Demand Forecasting Research Articles

Enhancing microgrid forecasting accuracy with SAQ-MTCLSTM: A self-adjusting quantized multi-task ConvLSTM for optimized solar power and load demand predictions

Accurate forecasting of solar power output and load demand is critical for the efficient operation and management of isolated microgrids, where reliability and sustainability are paramount. Traditional methods often struggle with data scarcity, limitations in capturing intricate temporal dynamics, and lack of scalability. This research introduces a novel multi-task learning (MTL) model, the Self-Aware Quantized Multi-Task ConvLSTM (SAQ-MTCLSTM), which addresses these challenges by jointly forecasting solar power and load demand while leveraging shared representations across these interdependent time series. The SAQ-MTCLSTM incorporates a sophisticated architecture that combines convolutional and LSTM layers with self-aware quantization to enhance computational efficiency and model adaptability. This allows for effective knowledge transfer, improved data utilization, and the capture of intricate temporal patterns. Evaluated on a real-world isolated microgrid dataset from the Micro Reseau Mafate research project, the model demonstrates significant improvements in forecasting accuracy, with MSE values of 0.0021 for solar power output and 0.0037 for load demand forecasting, compared to single-task and other advanced models. Extensive experiments highlight the impact of data scarcity, seasonality patterns, and microgrid topology on forecasting performance. Such forecasting is essential to optimize the integration and utilization of renewable resources, enhancing operational stability and reducing dependency on external energy supplies.

Techno-Economic Optimization of Microgrid Operation with Integration of Renewable Energy, Hydrogen Storage, and Micro Gas Turbine

Microgrids are integral to modern energy systems, yet they face substantial challenges in integrating diverse components, managing complex dynamics, and ensuring stability amid renewable energy variability. This study investigates the integration of wind turbines, an electrolyzer, and a hydrogen-compatible micro gas turbine (MGT), with a focus on enhancing operational efficiency and maintaining dynamic equilibrium within the microgrid. The synergy between the electrolyzer and MGT provides a robust energy storage solution, improving both system stability and performance.Using advanced machine learning and real operational data, this research generates highly accurate, rapid models with greater precision and detail than conventional methods. The intelligent management system developed combines real-time optimization and adaptive fine-tuning, using forecasts of weather, electricity prices, and energy demand to devise optimized operational strategies. The study systematically analyzes various configurations, including grid-connected and island modes, and evaluates the impact of hydrogen storage in each scenario.The results demonstrate that the optimization approach substantially enhances economic returns. However, this can also lead to increased natural gas consumption and emissions, revealing a trade-off between financial gains and environmental sustainability. This highlights the necessity for strategies that reconcile economic incentives with sustainability objectives, offering valuable insights for improved microgrid management.

INTEGRATING LAND USE ANALYSIS WITH WATER DEMAND ESTIMATION: A CASE STUDY OF PUTRAJAYA, MALAYSIA

The challenges posed by population growth, urbanization, and changing land use patterns on sustainable water resource management are significant. This paper puts forth an integrated framework aimed at assessing future water demand in Putrajaya, Malaysia. The proposed framework combines population projections, estimations of water demand, and analyses of land use activities. Through an examination of demographic trends and land use patterns, the framework predicts population growth and identifies areas with high water demand. Daily water use patterns in homes and businesses (temporal analysis) inform the designing future water infrastructure, incorporating temporal aspects. Statistical and spatial analysis techniques are then utilized to merge these projections with water demand estimations to quantify water requirements in various zones and types of land use. This study has unveiled two daily peaks in water demand, which align with household schedules. Residential areas emerge as the primary consumers of water, displaying an evening peak distinct from the midday peak seen in businesses. The current water demand in Putrajaya is estimated at 94 million litres per day, with domestic usage surpassing non-domestic usage in a ratio of 3:2. Projections based on future land use plans foresee a 19% increase in demand, underscoring the urgency for proactive water management strategies. Spatial analysis has highlighted residential areas as the main users of water, with demand levels varying throughout the city. By comprehending these temporal and spatial patterns, water authorities can strategically target interventions, optimize infrastructure siting, and forecast future demand trends. These proactive measures are essential for securing a sustainable water future for Putrajaya.

An Integrated Artificial Intelligence Approach for Building Energy Demand Forecasting

Buildings are complex assets, characterized by environments and uses that change over time, variable occupancies, and long life cycles. They have high operational costs, mostly due to their energy requirements, and account for 30% to 40% of global greenhouse gas emissions. Consequently, substantial effort has been made to forecast their energy needs, with the scope of optimizing their economic and environmental impact. In this regard, the available literature focuses mainly on short-term modeling through the implementation of sets of physics-based equations (i.e., white-box), functional relationships between input and output variables (i.e., black-box), or a combination of both (i.e., grey-box). On the other hand, more research is required on long-term forecast models with the aim of reducing the energy needs. Within this context, this article presents an original automatic procedure for forecasting the energy needs of buildings in short- and long-term time horizons. This is accomplished by scaling an unknown facility from a similar facility that is already known and by executing a black-box approach based on machine learning algorithms. The proposed method is implemented in real case studies in Italy, predicting the energy needs (i.e., heating, cooling, and electricity) of Sant’Anna Hospital in Ferrara using the historical data of Ca’ Foncello Hospital in Treviso. The results show an adjusted coefficient of determination above 0.7 and an average error below 10% for all the energy vectors, demonstrating a feasible forecast performance with a low training set-to-test set ratio.

Optimizing Distributed Energy Resources in Microgrid SCUC through Seq2Seq Scheduling Algorithms

In order to optimize Distributed Energy Resources (DERs) inside a microgrid's Security-Constrained Unit Commitment framework, this study investigates the use of Seq2Seq (Sequence-to-Sequence) scheduling methods (SCUC). There is a growing consensus that microgrids are an important part of the future of the electric grid because of the advantages they provide in terms of reliability, renewable energy integration, and overall efficiency. In the context of complicated SCUC issues, efficient scheduling and optimization of DERs are essential for realizing their full potential. Seq2Seq models, a kind of deep learning approach, have shown impressive performance in several applications requiring sequence prediction. Uncertainty in renewable energy production, energy demand forecasts, and security limitations are all addressed in this work as Seq2Seq algorithms are applied to microgrid SCUC. Significant gains in economic efficiency, ecological viability, and compliance with security constraints have been shown. This study lays the way for the development of more effective, sustainable, and resilient energy infrastructure by contributing to the advancement of the area of microgrid optimization.

Optimal inventory decisions of infrequently needed spare parts for maintenance with inaccurate demand forecasts and limited storage space

Abstract In modern manufacturing workshops, maintaining adequate inventory for spare parts is an effective strategy to satisfy maintenance requirements. However, due to the stochastic nature of failures, spare parts often have characteristics of infrequent usage and inaccurate forecasted demand. Additionally, the multitude diversity of spare parts and limited storage space in warehouse poses challenges to inventory decisions regarding the types and quantity of spares. Thus, this paper investigates the inventory decision optimization problem of infrequently needed spare parts for maintenance within limited storage space and inaccurate forecasted demand. This paper constructed an integer programming model based on chance constraints to address the uncertainty of the storage space occupied by the spare parts. The main contributions of this paper are as follows. First, we investigate the inventory decision optimization problem within limited storage space and inaccurate forecasted demand, which are rarely considered in the optimization problem of the spare parts for maintenance. Second, we proposed a method for simplifying the computation of chance constraints based on a transformation of the Central Limit Theorem. Third, we proposed a heuristic iterate algorithm to search the optimal inventory decision by minimizing the overall downtime and storage costs. Using data from the workshop of an automotive enterprise, a case study is conducted by comparison with the Genetic Algorithm.

Rethinking self-supervised learning for time series forecasting: A temporal perspective

Self-supervised learning has garnered significant attention for its ability to learn meaningful representations. Recent advancements have introduced self-supervised methods for time series forecasting. However, these efforts have faced limitations due to two primary drawbacks. Firstly, these approaches often borrow techniques from vision and language domains without adequately addressing the unique temporal dependencies inherent in time series data. Secondly, time series often show that the distribution shifts over time, which makes accurate forecasting challenging. In response to these issues, we propose TempSSL, a self-supervised learning framework designed for time series forecasting. TempSSL divides the time series data into context (history data) and target (future data), employing two pre-training strategies: (1) Temporal Masked Modeling (TMM) designed to capture temporal dependencies by reconstructing future time series based on historical context; (2) Temporal Contrastive Learning (TCL) employs context and target as positive samples to enhance discriminative representations and mitigate distribution shifts within the time series. TempSSL’s innovation lies in two key aspects. Firstly, it underscores the importance of temporal dependencies for time series forecasting by designing specific pre-training tasks. Secondly, it effectively integrates contrastive learning and masked modeling, leveraging their respective strengths to develop time series representation with strong instance discriminability and local perceptibility. Extensive experiments across seven widely used benchmark datasets demonstrate that TempSSL consistently outperforms existing self-supervised and end-to-end forecasting methods, achieving improvements ranging from 1.92% ∼ 78.12%. Additionally, TempSSL’s practical effectiveness is further demonstrated through successful application in natural gas demand forecasting.

Water Demand Forecasting in Multiple District Metered Areas Based on a Multi-Scale Correction Module Neural Network Architecture

Short-term water demand forecasting (STWDF) for multiple spatially and temporally correlated District Metering Areas (DMAs) is an essential foundation for achieving more refined management of urban water supply networks. However, due to the greater uncertainty associated with specific DMA demand compared to overall water usage, accurately predicting STWDF poses significant challenges. This study introduces an innovative network architecture—the multi-scale correction module neural network, built upon Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN) enhanced with Attention mechanisms—for simultaneous STWDF with a temporal resolution of one hour over a week for 10 DMAs located in a single city in northern Italy. This framework utilizes multivariate corrections to refine and enhance the output accuracy. The results reveal that, in comparison to traditional Gated Recurrent Unit or LSTM models, the proposed model with integrated correction modules, particularly those that leverage inter-DMA correlations, improves performance across all evaluation metrics by an average of 5%-20% per DMA. Additionally, it consistently delivers superior accuracy across three scenarios: single DMA forecasting, total water demand, and extreme conditions, while maintaining stable performance throughout. Furthermore, the interpretability analysis underscores the feasibility of this innovative structure and highlights the contribution of meteorological features to the predictive model in some DMA-level STWDF. The unified input-output framework elegantly simplifies the STWDF process across multiple DMAs, providing new insights and methodologies for future research in this domain.

Travel Demand Forecasting: A Fair AI Approach

Travel Demand Forecasting: A Fair AI Approach

MACHINE LEARNING APPROACHES FOR DEMAND FORECASTING: THE IMPACT OF CUSTOMER SATISFACTION ON PREDICTION ACCURACY

This study investigates the effectiveness of various machine learning models in predicting product demand based on customer satisfaction data. Four models—Linear Regression, Random Forest, Gradient Boosting, and Support Vector Machine (SVM)—were evaluated using performance metrics, including Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R² score. The results indicate that Gradient Boosting achieved the highest accuracy, with an MAE of 2.56, MSE of 12.75, RMSE of 3.57, and R² score of 0.82, effectively capturing the complex, non-linear relationships inherent in customer satisfaction factors. Random Forest also demonstrated strong performance, while Linear Regression and SVM showed limitations in handling intricate datasets. These findings underscore the importance of utilizing advanced machine learning techniques for accurate demand forecasting, highlighting the critical role of customer satisfaction data in enhancing predictive capabilities. The insights gained from this research can guide organizations in optimizing inventory management and improving customer satisfaction in a rapidly evolving market.

The integration of Artificial Intelligence in demand forecasting and inventory management in the United States

The integration of Artificial Intelligence in demand forecasting and inventory management in the United States has been examined in this paper. Demand forecasting and inventory management are two of critical areas in supply chain management, which is a veritable tool for promoting industrialization, manufacturing capabilities, and customers satisfaction. The use of AI in form of robotics, machine learning, deep learning, and predictive analytics, among others in all aspects of supply chain operations is gaining ground by the day. The integration of AI into the supply chain process can sustain multi-billion dollars trades in the United States, by reducing the cost of production and distribution, reducing human errors causing inaccurate demand forecasts, return shipment and cancellations of orders, etc. The challenges relating to the use of AI in demand forecasting and inventory management such as high cost of installation and maintenance, data privacy violations, requirement of skilled personnel, which are limited in global supply, and employees’ resistance to change were also identified. The outlook of relationship between Artificial Intelligence and supply chain management looks hopeful, brighter, and encouraging. This will be made possible by continuous development of AI capabilities and reducing the challenges of its widespread integration.

A critical review of machine learning applications in supply chain risk management

This article examines the transformative role of Machine Learning (ML) in Supply Chain Risk Management (SCRM), emphasizing its ability to enhance risk prediction and mitigation through real-time, data-driven insights. It demonstrates how ML applications like demand forecasting, inventory optimization, supplier risk assessment, and fraud detection improve supply chain resilience by analyzing large datasets to identify patterns, predict risks, and recommend proactive actions. However, the article also highlights key challenges, including data quality, availability, and privacy issues, which limit the effectiveness of ML models. Integrating ML with legacy systems poses additional technical and financial difficulties, particularly for smaller businesses. High implementation costs and scalability constraints further hinder widespread adoption. Ethical concerns, such as data bias and privacy, still need to be explored, raising questions about responsible ML use in supply chains. Future research should address these gaps by improving data governance frameworks for accuracy and privacy, developing scalable ML models suited to various supply chain environments, and exploring synergies with technologies like blockchain and the Internet of Things (IoT). These advancements will help realize ML’s full potential, fostering more agile, transparent, and resilient supply chains capable of navigating complex global risks. By tackling these challenges, ML can shift SCRM from reactive to proactive, ensuring long-term operational continuity and a competitive edge in the evolving global market.

Leveraging Artificial Intelligence for Optimizing Logistics Performance: A Comprehensive Review

Objective - Artificial Intelligence (AI) has become a pivotal technology in transforming logistics performance. This paper aims to comprehensively understand how AI-enabled solutions improve efficiency, accuracy, and responsiveness in logistics operations. The study focuses on synthesizing current research to explore AI applications across various logistics domains, such as predictive analytics, autonomous vehicles, and smart warehousing. Methodology/Technique – A systematic review approach was used to gather and analyze existing literature on AI applications in logistics. The review covered studies published in recent years, highlighting the advancements and impact of AI on logistics processes. The methodology included selecting relevant sources, categorizing AI applications, and assessing their effects on different logistics functions. Finding – The findings reveal that AI adoption substantially improves logistics operations, including enhanced operational performance, cost reduction, and increased customer satisfaction. Specific AI applications, such as predictive analytics for demand forecasting, autonomous vehicles for transportation, and smart warehousing for inventory management, were identified as key contributors to these improvements. However, challenges such as data privacy concerns and integration complexities were also noted. Novelty – This study's novelty lies in its comprehensive analysis of AI applications across various logistics domains, offering a holistic view of AI's role in optimizing logistics performance. This paper highlights the benefits of AI adoption and addresses the associated challenges, providing insights into future research directions and practical implications for leveraging AI in logistics. Type of Paper: Review JEL Classification: C61, C62, D83. Keywords: Artificial Intelligence; Logistics; Performance Improvement; Predictive Analytics; Autonomous Vehicles; Smart Warehousing Reference to this paper should be referred to as follows: Fatorachian, H. (2024). Leveraging Artificial Intelligence for Optimizing Logistics Performance: A Comprehensive Review, GATR-Global J. Bus. Soc. Sci. Review, 12(3), 146–160. https://doi.org/10.35609/gjbssr.2024.12.3(5) _____________________________________

Addressing cold start problems in new store locations with transfer learning in spatial GNNs

The cold start problem poses a significant challenge for retailers opening new store locations, primarily due to the lack of historical sales data necessary for accurate demand forecasting and effective inventory management. This paper explores the application of transfer learning within spatial Graph Neural Networks (GNNs) as a solution to this issue. By leveraging existing data from established stores that share similar characteristics, our proposed methodology enhances the forecasting accuracy and helps mitigate the risks associated with new store openings. We detail the architecture of the spatial GNN model, which captures complex spatial relationships and customer interactions, providing richer insights into demand patterns. Experimental results demonstrate substantial improvements in forecasting performance compared to traditional methods, highlighting the potential of transfer learning to inform strategic decision-making in retail. This research aims to provide actionable insights for retailers seeking to optimize their operations in new markets.

Effect of electricity policy uncertainty and carbon emission prices on electricity demand in China based on mixed-frequency data models

This paper first constructs the electricity policy uncertainty (EPU) in China with textual methods and analyses the effect of the EPU and carbon emission prices (CEPs) on the total electricity demand. This paper also forecasts the demand for electricity in China with three mixed-frequency data models. The results show that the EPU index efficiently captures the uncertainty of China's electricity policy. The effects of EPU and CEPs on electricity demand are significant, and incorporating them into the forecasting model will improve the accuracy and timeliness. Moreover, compared with the ARMA model and LSTM neural networks, mixed-frequency data models perform better in electricity demand forecasting.

The Integration of AI and Machine Learning in Supply Chain Optimization: Enhancing Efficiency and Reducing Costs

One of the biggest issues today is the increasing intricacy of supply chain networks and supply chain networks becoming more global. Following is the research paper on supply chain management accompanying the integration of AI & ML for effectiveness & efficiency & reduction of cost impacts. The purpose of the research is to assess the effectiveness of adoption of Artificial Intelligence and Machine Learning tools based on predictive analytics, automation, and real-time decision models with supply chain management tendencies in demand forecasting, inventory control, and logistics. In line with the research design that is mixed method, the data were obtained from high-impact case studies and industry reports with further support from the literature review. The usefulness analysis of AI and ML was conducted in line with the supply chain performance measures that include lead time, cost and system metrics for the actual implementation. The results suggest that the application of AI and ML leads to the key performance improvement in companies, such as an average of 20% decrease in operational costs and 15% shorter delivery times. The study will also provide a new understanding of the real world incorporation of AI and ML in supply chain and a path forward in the literature and practice. These technologies reveal the development prospect of how these supply chains can be rebuilt to be more robust as well as flexible.

Utilizing textual data from online reviews for daily tourism demand forecasting: A deep learning approach leveraging word embedding techniques

Accurately estimating daily tourism volumes is crucial for optimizing operational strategies and enhancing visitor experiences at tourist destinations. In this study, we leverage historical tourism volume data, search engine data, and online review data (including textual review contents) to forecast daily tourism demand. We develop a deep learning framework that includes a feature selection module to select search engine indices, a word embedding module to transform review texts into numerical predictors, and an extreme learning machine (ELM) enhanced with the whale optimization algorithm (WOA) for predictive modeling. Based on different word embedding techniques, we investigated two specific forecasting methods: one based on term frequency-inverse document frequency (TF-IDF) and Transformer, and the other based on the standard pre-trained bidirectional encoder representations from Transformers (BERT). These two methods enhance predictive accuracy while ensuring fast training and inference speeds, making it suitable for the high-frequency daily tourism forecasting task. Experimental results demonstrate that the two methods significantly outperform traditional approaches that rely solely on numerical data sources, achieving lower prediction errors and faster inference speeds compared to established benchmark methods, highlighting their potential to contribute to advancements in tourism prediction methodologies.

A Hybrid Approach for the Sales Forecasting of Paracetamol Products

The pharmaceutical industry is facing challenges due to various factors such as supply chain disruptions, changing consumer behavior, and regulatory changes. Accurate demand forecasting is essential to ensure an adequate supply of drugs. The goal of this work is to forecast paracetamol product demand. For this purpose, we propose a hybrid forecasting model combining two effective forecasting techniques: SARIMA (Seasonal AutoRegressive Integrated Moving Average) and ANFIS (Adaptive Neuro-Fuzzy Inference System). This proposal consists of nonlinear components of time series by ANFIS and adjusting the result by the mean of the residuals of the SARIMA to improve the accuracy and performance of ANFIS predictions. Before the prediction phase, we preprocess our data and detect the anomalies in our dataset with Locally Selective Combination in Parallel Outlier Ensembles (LSCP). Then, by treating these anomalies as missing values, they are imputed using the combination of Fuzzy-Possibilistic c-means (FCM) with support vector regression (SVR) and a genetic algorithm (GA). Finally, we evaluate the performance of the model and some known models based on MAPE. We choose the hybrid model SARIMA-ANFIS that provides the most accurate and reliable forecasting.

Optimizing the food supply chain through the integration of financial models and big data in procurement: A strategy for reducing food prices

Optimizing the food supply chain through the integration of financial models and big data analytics presents a transformative strategy for reducing food prices. In an increasingly complex global market, managing procurement effectively is crucial for cost control and price stabilization. Financial models, such as cost-benefit analysis, financial forecasting, and risk assessment, provide valuable insights into procurement decisions, helping organizations anticipate price fluctuations and manage expenses. When combined with big data analytics, these models enable more precise predictions and strategic planning. Big data analytics harnesses vast amounts of information from various sources, including market trends, supplier performance, and consumer behavior. By analyzing this data, organizations can identify patterns, optimize inventory levels, and forecast demand with greater accuracy. This data-driven approach enhances decision-making in procurement, allowing for better negotiation with suppliers, efficient inventory management, and the reduction of wastage. The integration of financial models with big data provides a holistic view of the supply chain, enabling companies to implement dynamic pricing strategies and adjust procurement practices based on real-time insights. This integration helps in minimizing operational costs, improving supply chain efficiency, and ultimately reducing food prices for consumers. For example, predictive analytics can identify potential supply disruptions and adjust procurement strategies proactively, while financial models can evaluate the cost-effectiveness of various sourcing options. Challenges such as data integration, quality, and privacy must be addressed to fully leverage these technologies. Moreover, investment in advanced analytics tools and skilled personnel is essential for successful implementation. In conclusion, the synergy between financial models and big data analytics offers a strategic approach to optimizing the food supply chain. By enhancing procurement practices and leveraging real-time insights, this strategy contributes to reducing food prices and improving overall supply chain efficiency, benefiting both businesses and consumers. Keywords: Food Supply Chain, Financial Models, Big Data Analytics, Procurement Optimization, Cost Control, Price Reduction, Inventory Management, Predictive Analytics, Supply Chain Efficiency.

Product Demand Forecasting For Inventory Management with Freight Transportation Services Index Using Advanced Neural Networks Algorithm

Purpose: Accurate demand forecasting is critical for optimizing inventory management, improving customer satisfaction, and maximizing profitability in the retail sector. Traditional forecasting models predominantly utilize micro-level variables, such as historical sales data and promotional activities, often neglecting the influence of macroeconomic conditions. Materials and Methods: This study addresses this gap by integrating Freight Transportation Services Index (TSI) which is an indicator for overall economic health with time series data of retail product sales. Utilizing an advanced neural networks model, we demonstrate that incorporating macroeconomic variables significantly enhances the model's predictive accuracy and explanatory power. Findings: The results reveal that the model with the TSI index outperforms conventional models, highlighting its potential for practical application in the industry. Implications to Theory, Practice and Policy: This approach offers a more comprehensive understanding of demand dynamics, enabling businesses to make more informed decisions, adapt to market fluctuations, and maintain a competitive edge.