Multiple Heterogeneous Sources Research Articles

Optimal scheduling of green heating systems in buildings considering economy and exergy efficiency

Green heating in buildings is of great significance to global energy conservation and emission reduction, and the design of optimal scheduling models is the key to achieve cost reduction and efficiency improvement in heating systems. For the building heating system containing multiple heterogeneous energy sources such as solar energy, air energy and electricity, a multi-objective optimal scheduling model is established with the goal of maximizing the exergy efficiency and minimizing the economic cost of the system, taking into account the quality difference of energy in the conversion process. Regarding the nonconvex and nonlinear constraints in the scheduling model, the incremental linearization method and Big-M method are used to linearize the transformation, and obtain the multi-objective mixed-integer linear programming model. Further, the ε-constraint method is used to solve the Pareto frontier of multi-objective optimization and the technique for order preference by similarity to ideal solution (TOPSIS) method is used to make decision. Finally, the simulation study of the model is carried out, and the results show that the proposed multi-objective optimal scheduling can balance the system operation economy and exergy efficiency compared with the belief rule base and evidential reasoning (BRB-ER) method, and the system’s daily economic cost are reduced by 9.66 %, 9.94 % and 17.24 %, respectively, and the system’s average daily coefficient of performance (COP) are improved by 4.66 %, 14.05 % and 13.22 %, respectively, on three typical days, it is verified that the proposed optimal scheduling model has better economy and exergy efficiency. The proposed multi-objective optimal scheduling model introduces exergy efficiency rather than energy efficiency as one of the optimization objectives of the heating system and reasonably solves the optimization model, which can provide a new idea and reference for the multi-objective optimal scheduling of the green heating system in buildings.

Embracing Multiheterogeneity and Privacy Security Simultaneously: A Dynamic Privacy-Aware Federated Reinforcement Learning Approach.

With growing demand for privacy-preserving reinforcement learning (RL) applications, federated RL (FRL) has emerged as a potential solution. However, existing FRL methods struggle with multiple sources of heterogeneity, while lacking robust privacy guarantees. In this study, we propose DPA-FedRL, the dynamic privacy-aware FRL framework, to simultaneously mitigate both issues. First, we innovatively put forward the concept of "multiheterogeneity" and embed the environmental heterogeneity into agents' state representations. Next, to ensure privacy during model aggregation, we incorporate a differentially private mechanism in form of Gaussian noise and modify its global sensitivity, tailored to suit FRL's unique characteristics. Encouragingly, our approach dynamically allocates privacy budget based on heterogeneity levels, which strikes a balance between privacy and utility. From the theoretical perspective, we give rigorous convergence, privacy, and sensitivity guarantees for our proposed method. Through extensive experiments on diverse datasets, we demonstrate that DPA-FedRL surpasses state-of-the-art approaches (PPO-DP-SGD, PAvg, and QAvg) in some highly heterogeneous environments. Notably, our novel privacy attack simulations enable quantitative privacy assessment, validating that DPA-FedRL offers over 1.359 × stronger protection than baselines.

Visual Intratumor Heterogeneity and Breast Tumor Progression.

High intratumoral heterogeneity is thought to be a poor prognostic indicator. However, the source of heterogeneity may also be important, as genomic heterogeneity is not always reflected in histologic or 'visual' heterogeneity. We aimed to develop a predictor of histologic heterogeneity and evaluate its association with outcomes and molecular heterogeneity. We used VGG16 to train an image classifier to identify unique, patient-specific visual features in 1655 breast tumors (5907 core images) from the Carolina Breast Cancer Study (CBCS). Extracted features for images, as well as the epithelial and stromal image components, were hierarchically clustered, and visual heterogeneity was defined as a greater distance between images from the same patient. We assessed the association between visual heterogeneity, clinical features, and DNA-based molecular heterogeneity using generalized linear models, and we used Cox models to estimate the association between visual heterogeneity and tumor recurrence. Basal-like and ER-negative tumors were more likely to have low visual heterogeneity, as were the tumors from younger and Black women. Less heterogeneous tumors had a higher risk of recurrence (hazard ratio = 1.62, 95% confidence interval = 1.22-2.16), and were more likely to come from patients whose tumors were comprised of only one subclone or had a TP53 mutation. Associations were similar regardless of whether the image was based on stroma, epithelium, or both. Histologic heterogeneity adds complementary information to commonly used molecular indicators, with low heterogeneity predicting worse outcomes. Future work integrating multiple sources of heterogeneity may provide a more comprehensive understanding of tumor progression.

Semantic Web Practices: Infrastructural Politics and the Future of the Web

Semantic Web Practices: Infrastructural Politics and the Future of the Web

Biparametric MRI for Local Staging of Prostate Cancer: Current Status and Future Applications.

Multiparametric magnetic resonance imaging (mpMRI) is the recommended modality for local staging of prostate cancer (PCa). The use of dynamic contrast-enhanced (DCE) imaging alone significantly improves staging performance. However, several studies have revealed that DCE imaging adds no extra benefit for PCa detection. Many authors observed benefits of performing prostate MRI without DCE, so called biparametric MRI (bpMRI), such as the elimination of the toxicity of gadolinium administration, reduction of examination time, costs and better accessibility. This narrative review describes the variety of imaging modalities in Local staging of PCa with bpMRI utilization and its comparison to mpMRI. A search of medical databases was performed to find eligible articles using the following key words: "prostate cancer", "MRI", "multiparametric", "biparametric" and "staging". MEDLINE, Web of Science, PubMed and Google Scholar were used to search for eligible articles published in the past 5 years and compared the diagnostic accuracy of mpMRI and bp MRI in local staging of PCa. A total of 48 articles were evaluated. Multiple systematic reviews used pooled data to compare the accuracy of biparametric and multiparametric examinations. However, all these studies advise caution on using pooled data for clinical practice, pointing to multiple sources of heterogeneity among the studies evaluated. Given the absence of prospective data comparing bpMRI and mpMRI, randomized, prospective, multicenter studies are encouraged. However, mpMRI is the recommended modality for local staging of PCa. It has superior performance compared to traditional staging based on clinical nomograms, and provides additional information on the site and extent of disease.

Research on the stability evolution mechanism and combinatorial optimization decision‐making of multitype heterogeneous energy cooperative operation

AbstractConsidering the cooperative willingness of multiple heterogeneous energy sources to participate in the alliance, the stability of the cooperative operation of each entity (hereinafter referred to as stability) and the supporting mechanism of its optimal combination to realize the high proportion of new energy generation are studied. Through simulating load peak‐cutting and valley‐filling in the power system, the difference between the load value and the base load in the 24 h of the daily load curve is used as the calculation condition for the supply balance of the alliance. Based on the source‐load power balance model, the feasibility of power data normalization is derived and demonstrated. Simultaneously, taking into account the time‐of‐use price cost of each entity, a multiscale operation cost allocation and stability evolution analysis model for the alliance is constructed using the Shapley value and the largest consistent set method. By comparing the willingness of each entity to engage in the alliance, as well as the multiscale operation stability and economic change rule of different alliances, the results demonstrate that: (1) Alliance 1 has the advantage of economic operation. (2) The common stability period of different alliances is influenced by the participation preferences of each entity (wind power, photovoltaic, and thermal power tend to participate in Alliance 1, while hydropower tends to participate in Alliance 2). (3) Alliance internal restructuring can be carried out by utilizing spatial and temporal difference characteristics of different entities’ participation in alliance preferences, thereby achieving alliance stability and efficient operation. This study provides a theoretical basis for making decisions regarding the optimized stable operation of the alliance.

Economic Evaluation of Neoadjuvant Versus Adjuvant Chemotherapy in Cancer Treatment: A Systematic Review and Meta-Analysis

ObjectivesIn the absence of evidence on whether neoadjuvant (NAC) or adjuvant chemotherapy (AC) is more beneficial for various tumor treatments, economic evaluation (EE) can assist medical decision making. There is limited evidence on their cost-effectiveness and their prospective evaluation is less likely in the future. Therefore, a systematic review and meta-analysis about EE for NAC versus AC in solid tumor help compare these therapies from various perspectives. MethodsVarious databases were searched for studies published from inception to 2021. This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses reporting guidelines and economic-specific guidelines. The data were pooled using a random effects model when possible. ResultsThe retrieval identified 15 EE studies of NAC versus AC in 8 types of cancer. NAC is the dominant strategy for pancreatic, head and neck, rectal, prostate cancers and colorectal liver metastases. For ovarian cancer, NAC is cost-effective with a lower cost and higher or similar quality-adjusted life-year. There were no significant differences in cost and outcomes for lung cancer. For stage IV or high-risk patients with ovarian or prostate cancer, NAC was cost-effective but not for patients who were not high risk. ConclusionsThe EEs results for NAC versus AC were inconsistent because of their different model structures, assumptions, cost inclusions, and a shortage of studies. There are multiple sources of heterogeneity across EEs evidence synthesis. More high-quality EE studies on NAC versus AC in initial cancer treatment are necessary.

Multi-source and multimodal data fusion for improved management of a wastewater treatment plant

Wastewater treatment plants (WWTPs) are complex systems presenting stochastic, non-linear, and non-stationary behavior, which makes their operational management very challenging. In this context, data collected from distributed sources across the plant play a central role in the optimized operation and control of WWTPs. However, even when available, the use of collected data is far from trivial due to the coexistence of asynchronous measurements, data with different granularity, measurements of different quality (precision, accuracy), multimodal sources (sensors, spectra, images, hyphenated instrumentation), among other aspects related to the data life cycle. Such heterogeneity in process data characteristics hinders the application of most off-the-shelf data analytics methods. Flexible solutions able to cope with the complexity of systems and of the data they generate are therefore necessary to overcome these limitations and enable an effective analysis and operation of WWTPs. In this article, data-fusion approaches for handling multiple heterogeneous sources of process data are developed and comparatively tested. Priority is given to solutions that can flexibly be adapted to different specific operational contexts. The methodologies are tested on an industrial case study (WWTP), where the concentration of a toxin in the effluent stream is to be predicted from available heterogeneous data. Single- and multi-source modeling approaches are contemplated and a nested cross-validation method was developed to handle the time-series nature of the models. Bayesian fusion synergistically combines data from different sources considering their uncertainty, standing out among the methodologies tested as offering a good balance in terms of accuracy (RMSEP = 1.34), stability (prequential IQR = 0.034), and flexibility (to accommodate missing and new sources).

Schema matching based on energy domain pre-trained language model

Data integration in the energy sector, which refers to the process of combining and harmonizing data from multiple heterogeneous sources, is becoming increasingly difficult due to the growing volume of heterogeneous data. Schema matching plays a crucial role in this process by giving each representation a unique identity by matching raw energy data to a generic data model. This study uses an energy domain language model to automate schema matching, reducing manual effort in integrating heterogeneous data. We developed two energy domain language models, Energy BERT and Energy Sentence Bert, and trained them using an open-source scientific corpus. The comparison of the developed models with the baseline model using real-life energy domain data shows that Energy BERT and Energy Sentence Bert models significantly improve the accuracy of schema matching.

Mobilising Long-Term Natural Environment and Biodiversity Data and Exposing it for Federated, Semantic Queries

Biodiversity and ecosystems cannot be studied without assessing the impacts of changing environmental conditions. Since the 1980s, the U.S. National Science Foundation's Long Term Ecological Research (LTER) Network has been a major force in the field of ecology to better understand ecosystems. In Europe, the LTER developments are led by the the Integrated European Long-Term Ecosystem, critical zone and socio-ecological system Research Infrastructure (eLTER RI), a currently project-based infrastructure initiative with the aim to facilitate high impact research and catalyse new insights about the compounded impacts of climate change, biodiversity loss, soil degradation, pollution, and unsustainable resource use on a range of European ecosystems and socio-ecological systems. The European LTER network, which forms the basis for the up-coming eLTER RI, is active in 26 countries and has 500 registered sites that provide legacy data e.g., historical time-series data about the environment (not only biodiversity). Its site information and dataset metadata with the measured variables are available to be searched at the Dynamic Ecological Information Management System - Site and dataset registry (DEIMS-SDR, Wohner et al. 2019). While DEIMS-SDR data models utilize parts of the Ecological Metadata Language (EML) schema 2.0.0, location information follows the European INSPIRE specification. The future eLTER data is planned to consist of site-based, long-term time-series of ecological data. The eLTER projects have defined eLTER Standard Observations (SO), which will include the minimum set of variables as well as the associated method protocols that can characterise adequately the state and future trends of the Earth's systems. (Masó et al. 2020, Reyers et al. 2017). The current eLTER network consists of sites that differ in terms of infrastructure maturity or environment type and may focus on one or several of the future SOs or they are not yet executing any holistic monitoring scheme. The main objective is to convert the eLTER site network into a distributed research infrastructure that incorporates a clearly outlined mandatory monitoring program. Essential to this effort are the suggested variables for eLTER SOs and the corresponding methods and protocols for relevant habitat types according to the European Nature Information System (EUNIS) in each domain. eLTER variables are described by using the eLTER thesaurus "EnvThes". These descriptions are currently enhanced by the use of the InteroperAble Descriptions of Observable Property Terminology (I-ADOPT, Magagna et al. 2022) framework to provide the necessary level of detail required for seamless data discovery and integration. Variables and their associated methods and protocols will be formalised to enable automatic site classifications, by building on existing observation representations such as the Extensible Observation Ontology (OBOE), Open Geospatial Consortium's Observation and Measurement, and the future eLTER Standard Observation ontology. DEIMS-SDR will continue to be used as a core service with an RDF representation of its assets (sites, sensors, activities, people) currently being implemented. This action is synced with the Biodiversity Digital Twin (BioDT) project to ensure maximum findability, accessibility, interoperability and re-usability (FAIRness; Wilkinson et al. 2016) of data through FAIR Digital Objects (FDO). Other (digital) assets such as datasets, models and analytical workflows will be documented in the Digital Asset Register (DAR) alongside semantic mapping and crosswalk techniques, to provide machine-actionable metadata (Schultes and Wittenburg 2019, Schwardmann 2020). The Biodiversity Digital Twin (BioDT) project is bringing together biodiversity and natural environment data from seven thematic use cases for modeling. BioDT prototypes rely on openly available data that comes from multiple heterogeneous sources using a multitude of standards and formats. In the pilot phase, merging data requires "hand picking" from selected sources, and automation of workflows would still require many additional steps. There are ongoing efforts in both the BioDT and eLTER projects to find best ways and practices to bring the raw data together by using suitable standards but also to harmonise the other environment variables by referring to vocabularies and possibly express the data as FDOs. Currently both the EML schema and Darwin Core standard (Darwin Core Task Group 2009; with registered extensions) allow referring to external schemas and vocabularies, which give flexibility but may still prove to be too narrow for the multitude of data types and formats the natural environment data requires. We welcome discussion about how to create good practices for enriching and harmonising natural environment data and species occurrence data in a meaningful way. GBIF's new data model and enriching the raw data with semantic artefacts may prove to be the way to provide thematic data products that combine data from multiple sources.

A Correlation Analysis and Visualization of Climate Change using Post-Disaster Heterogeneous Datasets

<p>There are numerous geo-climatic and human factors that contribute to the occurrence of natural disasters in the real-world scenario. Besides the study of causes and preconditions of such calamities, post-disaster analysis is essential for the efficient management of the disaster situation. This process needs timely and accurate data in light of the increasing frequency and severity of climate change-related extreme weather events. The analysis of disaster data involves the challenging task of integrating multiple heterogeneous sources, data ingestion and visualization. This paper aims at providing a three-dimensional analytical view of disaster data as time-series charts and a statistical model to evaluate the correlation between the occurrence of disasters, climate change and the corresponding economic damages (as a percentage of GDP). The created corpus consists of two unified heterogeneous datasets. The first one forms the basis for exploratory data analysis involving visualization charts for disaster-related factors. The second dataset involves temperature anomaly information which allows for climate change analysis. The integration of the above datasets provides scope for correlation analysis using three different variations of coefficients. Therefore, statistical methodologies are leveraged to play important role in managing disasters, from preparation to recovery and reporting. The graphical representations provide insights on regional trends that follow, related to factors such as the proportion of each type of disaster in the various losses incurred. Therefore, obtaining reliable information about the population, the economy and climate are crucial both for risk management and preparedness and for responding to disasters. The research puts forth a detailed statistical methodology with a spatial dimension to study the impacts on the economy and infrastructure in the aftermath of a disaster thereby ensuring specialized assistance. The inference of the analysis confirmed a positive correlation between climate change and occurrence of natural disasters. Therefore, statistical evidence of an important phenomenon like climate change affecting natural disasters brings awareness among the population in the society to be more environmentally responsible.</p>

Estimating the Optimal Inflation Target from Trends in Relative Prices

We derive closed-form expressions for the optimal inflation target under Calvo and menu-cost frictions in a model featuring a product life cycle and multiple sources of heterogeneity across goods. We show that both frictions deliver the same optimal target and how it can be estimated from observed trends in relative prices. Using micro price data underlying the CPI in the United Kingdom, we document that relative prices fall over the product lifetime in most expenditure items. This causes the optimal inflation target to be positive, with our baseline estimate for the United Kingdom being equal to 2.6 percent in 2016. The optimal target has steadily increased over the prior two decades due to an acceleration in the rates of relative price decline over time. (JEL C51, D15, E31, L11)

The Brain's Control Networks in Reading: Insights From Cross-Task Studies of Youth.

Humans engage multiple brain systems to read successfully, including using regions important for vision, language, and control. Control refers to the set of executive processes in the brain that guide moment-to-moment behavior in service of our goals. There is a growing appreciation for the role of the brain's control system in reading comprehension, in reading skill change over time, and in those who have difficulty with the reading process. One way to understand the brain's control engagement in reading may be to study control engagement across multiple tasks in order to study consistencies, or cross-task similarities, relative to reading-specific variations. In this commentary, I briefly summarize some of our recent work studying the brain's control networks across different tasks (e.g., when reading, or doing different executive function tasks). I then review our findings of when control activation does or does not relate to measures of reading ability, and reading growth over time. The utility of cross-task comparisons in neuroimaging is noted, as well as the need to better understand multiple sources of heterogeneity in our developmental samples. I end by discussing a few of the many future directions for further study of the brain with regard to the brain's control processing and academic achievement.

Federated Generative Model on Multi-Source Heterogeneous Data in IoT

The study of generative models is a promising branch of deep learning techniques, which has been successfully applied to different scenarios, such as Artificial Intelligence and the Internet of Things. While in most of the existing works, the generative models are realized as a centralized structure, raising the threats of security and privacy and the overburden of communication costs. Rare efforts have been committed to investigating distributed generative models, especially when the training data comes from multiple heterogeneous sources under realistic IoT settings. In this paper, to handle this challenging problem, we design a federated generative model framework that can learn a powerful generator for the hierarchical IoT systems. Particularly, our generative model framework can solve the problem of distributed data generation on multi-source heterogeneous data in two scenarios, i.e., feature related scenario and label related scenario. In addition, in our federated generative models, we develop a synchronous and an asynchronous updating methods to satisfy different application requirements. Extensive experiments on a simulated dataset and multiple real datasets are conducted to evaluate the data generation performance of our proposed generative models through comparison with the state-of-the-arts.

A bi-level scheduling strategy for integrated energy systems considering integrated demand response and energy storage co-optimization

In the context of China's “double carbon” commitment to the world, the introduction of integrated demand response mechanism and compressed air energy storage system into the traditional energy system is important to improve its structure, promote the interaction of multiple heterogeneous energy sources, and improve energy conversion efficiency. As such, in this study, we constructed an upper-layer capacity configuration model aiming to optimize the operating costs of compressed air energy storage systems during the planning period and subsequently integrated a heterogeneous energy network operation and China's certified emission reduction constraints under various operating scenarios. We established a lower-level optimal scheduling model with the optimization objective of minimizing the energy purchase, operation, maintenance, integrated demand response subsidy, and carbon trading costs. Finally, the KKT condition and the Big M method were used to solve this two-tier optimization problem. The comparative analysis of the three schemes showed that, under the consideration of the integrated demand response mechanism and the equipped compressed air energy storage system, the total operating cost of the system was reduced by an average of 2.28 %, the carbon emission was reduced by an average of 7.17 %, the profit from the sale of surplus carbon emission credits increased by an average of 16.7 %, and the energy storage capacity was reduced by approximately 39 %. Thus, this method effectively reduced the capacity of the energy storage system and significantly improved its reliable, economic, and low-carbon operation.

Day-ahead scheduling of the electrothermal integrated system considering multiple thermal inertia

For the problem of optimal scheduling of integrated energy systems involving wind power consumption, by mining multiple thermal inertia, a stochastic optimal scheduling strategy with consideration of mixed uncertainties is proposed. Firstly, the network thermal inertia is explored by considering the dynamic characteristics and thermal storage of heat network. Then the heat load flexibility constraint is introduced to model the thermal inertia of buildings. Finally heat pumps and heat storage are used as time-shiftable loads or sources, which are jointly modelled with CHP units to solve the peak-valley mismatch of electric and heat loads, and on this basis, the model is established. In this paper, the Barry Island’s case which is composed of 33-node thermal network coupled with 9-node grid system is used as an example for analysis, and results show that by exploring the multiple thermal inertia, both the complementarity and the flexible dispatch between multiple heterogeneous energy sources in integrated energy system can be realized.

UFODMV: Unsupervised Feature Selection for Online Dynamic Multi-Views

In most machine learning (ML) applications, data that arrive from heterogeneous views (i.e., multiple heterogeneous sources of data) are more likely to provide complementary information than does a single view. Hence, these are known as multi-view data. In real-world applications, such as web clustering, data arrive from diverse groups (i.e., sets of features) and therefore have heterogeneous properties. Each feature group is referred to as a particular view. Although multi-view learning provides complementary information for machine learning algorithms, it results in high dimensionality. However, to reduce the dimensionality, feature selection is an efficient method that can be used to select only the representative features of the views so to reduce the dimensionality. In this paper, an unsupervised feature selection for online dynamic multi-views (UFODMV) is developed, which is a novel and efficient mechanism for the dynamic selection of features from multi-views in an unsupervised stream. UFODMV consists of a clustering-based feature selection mechanism enabling the dynamic selection of representative features and a merging process whereby both features and views are received incrementally in a streamed fashion over time. The experimental evaluation demonstrates that the UFODMV model has the best classification accuracy with values of 20% and 50% compared with well-known single-view and multi-view unsupervised feature selection methods, namely OMVFS, USSSF, and SPEC.

Predicting mechanically ventilated patients future respiratory system elastance – A stochastic modelling approach

Background and objectiveRespiratory mechanics of mechanically ventilated patients evolve significantly with time, disease state and mechanical ventilation (MV) treatment. Existing deterministic data prediction methods fail to comprehensively describe the multiple sources of heterogeneity of biological systems. This research presents two respiratory mechanics stochastic models with increased prediction accuracy and range, offering improved clinical utility in MV treatment. MethodsTwo stochastic models (SM2 and SM3) were developed using retrospective patient respiratory elastance (Ers) from two clinical cohorts which were averaged over time intervals of 10 and 30 min respectively. A stochastic model from a previous study (SM1) was used to benchmark performance. The stochastic models were clinically validated on an independent retrospective clinical cohort of 14 patients. Differences in predictive ability were evaluated using the difference in percentile lines and cumulative distribution density (CDD) curves. ResultsClinical validation shows all three models captured more than 98% (median) of future Ers data within the 5th – 95th percentile range. Comparisons of stochastic model percentile lines reported a maximum mean absolute percentage difference of 5.2%. The absolute differences of CDD curves were less than 0.25 in the ranges of 5 < Ers (cmH2O/L) < 85, suggesting similar predictive capabilities within this clinically relevant Ers range. ConclusionThe new stochastic models significantly improve prediction, clinical utility, and thus feasibility for synchronisation with clinical interventions. Paired with other MV protocols, the stochastic models developed can potentially form part of decision support systems, providing guided, personalised, and safe MV treatment.

Heterogeneity of Zika virus exposure and outcome ascertainment across cohorts of pregnant women, their infants and their children: a metadata survey

ObjectivesTo support the Zika virus (ZIKV) Individual Participant Data (IPD) Consortium’s efforts to harmonise and analyse IPD from ZIKV-related prospective cohort studies and surveillance-based studies of pregnant women and their...

Bi-level optimal scheduling of virtual energy station based on equal exergy replacement mechanism

Integrated Demand Response (IDR) is an effective way to make full use of the resources of the integrated energy system (IES) to achieve a positive interaction between supply and demand. Energy cascade utilization is one of the important technical principles of the IES to improve energy utilization efficiency and realize comprehensive benefit optimization. The design of an incentive mechanism is an important prerequisite for realizing IDR. When formulating the incentive mechanism, the coupling between the conversion efficiency of multiple energy sources and the depreciation of energy quality in the process of energy cascade utilization should be comprehensively considered. The exergy is the index that can evaluate the engineering value of any form of energy. Therefore, this paper unifies the multiple heterogeneous energy sources in the IES in the form of exergy through the energy quality coefficient. To more accurately reflect the difference in energy value of multiple heterogeneous energy sources during the time of invocation, an equal exergy replacement mechanism is designed according to the principle of high energy quality and high price. The regulated energy of the customer participating in the IDR is stored in the form of exergy, and the customer can exchange any form of energy at any time under the same exergy. On this basis, a bi-level optimal scheduling model for virtual energy station (VES) to participate in the day-ahead market is established, which realizes the improvement of multi-subject interests. And use Information Gap Decision Theory (IGDT) to deal with the risks posed by uncertainty in day-ahead market prices. Finally, the case study shows that the designed incentive mechanism can further reduce the energy purchase cost of customers, and VES can obtain higher benefits than only providing internal energy sales services. The IGDT method can reduce operational risk, which verifies the feasibility of the method.