Demand Data Research Articles

Multi-Objective Distributionally Robust Optimization for Earthquake Shelter Planning Under Demand Uncertainties

Deciding the locations of shelters and how to assign evacuees to these locations is crucial for effective disaster management. However, the inherent uncertainty in evacuation demand makes it challenging to make optimal decisions. Traditional stochastic or robust optimization models tend to be either too aggressive or overly conservative, failing to strike a balance between risk reduction and cost. In response to these challenges, this research proposes a multi-objective distributionally robust optimization (MODRO) model tailored for shelter location and evacuation allocation. First, an ambiguity set (moment-based or distance-based) is constructed to capture the uncertainty in evacuation demand, reflecting the possible range of outcomes based on demand data from a disaster simulation model. Then, the distributionally robust optimization model considers the “worst-case” distribution within this ambiguity set to minimize construction cost, travel distance, and unmet demand/unused capacity, balancing the trade-off between overly conservative and overly optimistic approaches. The model aims to ensure that shelters are optimally located and evacuees are efficiently allocated, even under the most challenging scenarios. Furthermore, Pareto optimal solutions are obtained using the augmented ε-constraint method. Finally, a case study of Ogu, a wooden density built-up area in Tokyo, Japan, compares the DRO model with stochastic and robust optimization models, demonstrating that the cost obtained by the DRO model is higher than a stochastic model while lower than the worst-case robust model, indicating a more balanced approach to managing uncertainty. This research provides a practical and effective framework for improving disaster preparedness and response, contributing to the resilience and safety of urban populations in earthquake-prone areas.

Intelligent fuzzy-PSO based grid connected solar EV charging station for electric buses to enhance stability and energy management

Abstract Modern power networks are relying more and more on Distributed Generation (DG) and solar energy-powered Electric Bus (EB) charging stations; nevertheless, controlling the charging process under situations of peak EB load and variable solar irradiation presents considerable hurdles. This study presents a coordinated control approach that maximizes energy distribution for solar (PV)-based EB charging stations by using an intelligent energy management system (IEMS). Through the analysis of load demand and meteorological data in real time, the IEMS optimizes PV generation and grid power consumption, thereby halving peak power demand, minimizing system losses, and easing distribution grid stress. In order to enable dynamic modifications to the energy flow between PV power, the grid, and battery storage, the method integrates an adaptive neuro-based fuzzy control system that anticipates solar energy generation and predicts EB load demands. Furthermore, to balance power distribution between battery and ultracapacitor systems and control energy storage in ultracapacitors, a fuzzy inference system optimized using Particle Swarm Optimization (PSO) is utilized. This prolongs battery life and guarantees effective energy management. The overall performance of the IEMS is improved by the fuzzy logic controller, which produces output signals that specify the best power distribution for any energy storage system. Digital simulations and a real-time hardware-in-loop experimental setup are used to validate the effectiveness of the proposed system, which shows notable gains in energy management, grid stability, and system efficiency. In order to improve intelligent energy management in grid-connected solar-powered systems, this study offers a viable method for incorporating renewable energy sources into EB charging stations.

Guiding transdisciplinary synthesis processes for social-ecological policy decisions

Interdisciplinary synthesis research has been promoting significant advances in expanding academic knowledge. However, its application to address social-ecological problems poses challenges, typical of transdisciplinary research and co-production initiatives. Based on the experience of seven working groups from a Brazilian synthesis nucleus dedicated to co-producing social-ecological public policies, we present eight learnings to strengthen transdisciplinary syntheses. Those syntheses require flexibility in the working group dynamics to facilitate collaborative work, with frequent and short meetings held in easily accessible locations (1) for the involved stakeholders. They also require flexibility to shape different trajectories, depending on demand urgency, data and knowledge availability (2). Flexibility is also required to adjust to political circumstances, acknowledging that there are trade-offs between responding to urgent political needs and creating novel ideas, knowledge and outputs (3). In addition, the creation of formal institutions, particularly, formal engagement at the science-policy interface (4) and creating formal platforms for disseminating non-academic outputs (5) are key to stimulate the involvement of policy-makers and scientists in collaborative transdisciplinary syntheses. Symmetrical, horizontal interactions within a two-way science-policy linkage (6), alongside collective reflexivity on bridging diverse knowledge, skills, and authorities (7) are crucial for aligning academic knowledge with policy practices. Active involvement of individuals skilled in both scientific research and policy-making, who act as knowledge brokers, further strengthens this alignment. Finally, attention to create positive interactions and transparently communicating help to build trust among participants (8). These adjustments can enhance the potential of transdisciplinary syntheses to generate actionable knowledge at the science-policy interface.

A switching based forecasting approach for forecasting sales data in supply chains

Forecasting future demand has been a challenging task for supply chain practitioners, which is further exacerbated due to the recent pandemic effects. While the literature suggests a potential for improved accuracy with ML/AI approaches compared to probabilistic distribution-based traditional forecasting methods, the extent of this enhancement may vary based on the specific case. It is recognized that traditional probabilistic forecasting approaches are often considered less accurate and may lead to errors, potentially influencing the estimation of overall business costs. Meanwhile, with the advancement of artificial intelligence (AI) approaches, such as machine learning (ML) and deep learning (DL), this misestimation of cost can be reduced by forecasting demand more accurately from historical data. Consequently, this paper applies several AI-based approaches to predict demand data. Since no fixed AI approach works best for all datasets, a switching-based forecasting approach (SBFA) is proposed to exploit the merit of different advanced ML/DL approaches for different days ahead of prediction. Based on the performance of validation data, the proposed system automatically switches between different approaches to determine a more appropriate forecasting approach. A two-echelon supply chain model with different attributes is developed to validate the proposed SBFA against a few traditional forecasting approaches. The reorder points of this supply chain model are calculated based on the predictions from conventional/ML/DL forecasting approaches. Predictions from SBFA and other approaches are analysed by calculating overall supply chain cost. Based on overall supply chain costs under static and dynamic lead time settings, the effectiveness and applicability of the proposed SBFA against traditional forecasting approaches are demonstrated.

A hybrid deep learning framework for short-term load forecasting with improved data cleansing and preprocessing techniques

The dynamic variations in connected loads and the intermittent nature of renewable energy sources significantly impact smart grid reliability. Accurate load and generation forecasting stand pivotal for enhancing grid reliability and efficient operations. In this study, a comprehensive four-step approach is introduced for short-term load forecasting (STLF) aimed at precisely estimating power demand and generation. The process unfolds with data collection, followed by rigorous standardization, preprocessing, and cleansing of demand and generation data. Subsequently, a hybrid deep learning model, comprising bidirectional long short-term memory (BiLSTM), bidirectional gated recurrent unit (BiGRU), and a fully connected layer is trained using the clean data. This model harnesses the temporal dependencies within the data for accurate predictions. The model's performance is then evaluated using mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE), providing forecasted values for both generation and demand on minute, hourly, daily, and weekly intervals. Notably, the proposed approach achieves a remarkable MSE of 0.0058 for load forecasting and 0.0033 for generation forecasting. Comparative analysis with state-of-the-art (SOTA) techniques in terms of accuracy and computational cost underscores the superior accuracy of the proposed framework in forecasting both generation and demand. Importantly, the proposed approach bridges the gap in reliability enhancement for smart grids operating, a facet lacking in many existing methodologies. This signifies the potential of the proposed approach to bolster smart grid reliability, ensuring more reliable operations.

The scope of demand-driven material requirements planning in operative purchasing of a multi-national company: A case study

The objective of this study is to address existing study gaps by defining what materials are demand-driven material requirements planning (DDMRP) suitable and building a tool that helps to identify such materials. The research problems are approached with three different questions about suitability, features of the identification tool, and financial impact. The research methodology consists of a mixed method case study approach, where semi-structured interviews were conducted with supply chain professionals from the case company, and quantitative data related to the case company’s operations was analyzed. The data consisted of the relevant data of over 10,000 purchased materials. The findings of this study suggest that there are no certain characteristics that materials suitable for DDMRP have, but the potential must be defined individually in the case of every purchased material. The tool initially developed in this study helps to identify materials that meet the requirement of providing potential positive financial impact if brought into DDMRP scope by analyzing historical demand data, lead times, and inventory carrying costs.

Transformer-based probabilistic demand forecasting with adaptive online learning

Demand forecasting is crucial for the operation and planning of the energy and power industry. Accurate demand forecasting can assist decision-makers in reducing operational or planning costs while optimizing supply chains and resource utilization. Due to various uncertainties and dynamically changing environments, traditional offline deterministic forecasting methods may face difficulty in forecasting demands accurately and be unable to effectively quantify the uncertainties in forecasts. To adapt to dynamic environmental changes, several online learning methods have been proposed, which require large computational resources. To overcome this limitation, this paper proposes a Transformer-based probabilistic demand forecasting with an adaptive online learning algorithm. More specifically, we design a probabilistic decoder for the Transformer to quantify forecast uncertainty. Furthermore, an adaptive online learning algorithm is employed to selectively update parameters based on an adaptive update strategy, reducing computational burden while allowing for timely adaptation to dynamic environmental changes. The performance of the proposed method is evaluated on multiple benchmark datasets and real electricity demand data from fused magnesium furnace (FMF). The results show that the proposed method exhibits superior accuracy in both deterministic and probabilistic forecasting scenarios. Finally, through ablation experiments, the effectiveness of the proposed adaptive online probabilistic forecasting algorithm is validated.

The multi-depot pickup and delivery vehicle routing problem with time windows and dynamic demands

The rapid development of the urban logistics recycling industry, combined with the complexity of the pickup and delivery networks, has created a surge in dynamic customer demands and exacerbated the difficulty of logistics resource sharing. Accordingly, this work focuses on a multi-depot pickup and delivery vehicle routing problem with time windows and dynamic demands, which incorporates resource sharing. A bi-objective mathematical model is formulated to minimize the total operating cost and number of vehicles. A three-dimensional affinity propagation clustering and an adaptive nondominated sorting genetic algorithm-II are combined to find Pareto optimal solutions. A dynamic demand insertion strategy is proposed to determine the vehicle service sequences for dynamic situations. Combined with an elite iteration mechanism to prevent the proposed algorithm from falling into search stagnation and improve the convergence performance. The superiority of the proposed algorithm is verified by comparing with CPLEX solver (i.e., ILOG CPLEX Optimization Studio 12.10), multi-objective ant colony optimization, multi-objective particle swarm optimization, multi-objective evolutionary algorithm, multi-objective genetic algorithm, and decomposition-based multi-objective evolutionary algorithm with tabu search. Besides, the proposed model and algorithm are tested by a real-world case study in Chongqing city, China, and the further analysis indicates that significant improvement can be achieved. Furthermore, by incorporating the recognition and prediction techniques of artificial intelligence on dynamic demand data, the proposed approach can realize the self-optimization of multi-depot vehicle routes and the precise allocation of logistics resources in dynamic environments. This study is conducive to the construction of a digitally-intelligent urban logistics system.

Citrus greening, retail fruit prices, and household welfare: A demand system approach

Abstract Pests and diseases like citrus greening that threaten agricultural productivity also pose a risk to consumers. Reductions in food supply due to outbreaks and spread could increase food prices. We model U.S. household fruit demand using a Quadratic Almost Ideal Demand System and data from Circana’s 2020 and 2021 household panel. Price and income elasticity estimates reveal how household behavior might adjust with shocks to citrus and other fruit prices. Shocks to retail fruit prices can be from either citrus greening or other phenomena such as adverse weather. We also use compensating variation to estimate the impact that changes in fruit prices could have on consumer welfare.

Code and Data Repository for A Data-Driven Optimization Framework for Static Rebalancing Operations in Bike Sharing Systems

This repository provides an end-to-end solution to the static rebalancing operations in bike sharing systems. The study uses the raw data of bike demand and station inventory status to generate the optimal rebalancing routes that reallocate system-wide bike inventories among stations during the night to maintain a high service level while minimizing demand loss due to stockout or overcapacity. The repository reports the raw data, algorithms, and extensive numerical experiments reported in the paper, using real-world data from New York City Citi Bike.

Explainable artificial intelligence for reliable water demand forecasting to increase trust in predictions

The “EU Artificial Intelligence Act” sets a framework for the implementation of artificial intelligence (AI) in Europe. As a legal assessment reveals, AI applications in water supply systems are categorised as high-risk AI if a failure in the AI application results in a significant impact on physical infrastructure or supply reliability. The use case of water demand forecasts with AI for automatic tank operation is for example categorised as high-risk AI and must fulfil specific requirements regarding model transparency (traceability, explainability) and technical robustness (accuracy, reliability). To this end, six widely established machine learning models, including both transparent and opaque models, are applied to different datasets for daily water demand forecasting and the requirements regarding model accuracy, transparency and technical robustness are systematically evaluated for this use case. Opaque models generally achieve higher prediction accuracy compared to transparent models due to their ability to capture the complex relationship between parameters like for example weather data and water demand. However, this also makes them vulnerable to deviations and irregularities in weather forecasts and historical water demand. In contrast, transparent models rely mainly on historical water demand data for the utilised dataset and are less influenced by weather data, making them more robust against various data irregularities. In summary, both transparent and opaque models can fulfil the requirements regarding explainability but differ in their level of transparency and robustness to input errors. The choice of model depends also on the operator's preferences and the context of the application.

A multi-objective production scheduling model and dynamic dispatching rules for unrelated parallel machines with sequence-dependent set-up times

This study presents a production scheduling for unrelated parallel machines with machine and job sequence-dependent setup times. The system performance measures to minimize include makespan, total tardiness, and number of tardy jobs. The aim of the study is to develop a solution methodology that can solve the problem for the large scale. First, the problem is formulated as a mixed-integer linear programming model. The augmented ɛ-constraint method is applied to find Pareto solutions for small problem instances. The purpose is to demonstrate that Pareto solutions, which balance the trade-offs among three measures of performance, can be found for these instances. Dispatching rule-based heuristics is developed to solve large problem instances. The heuristics feature three dispatching rules that are designed to handle the dependent setup time. In addition, these rules are combined into six variants using a time-based rule-switching mechanism. The heuristics are tested with 18 problem instances, containing 244 to 298 jobs, in two demand scenarios derived from the monthly demand data from an industrial user. Under each demand scenario, a set of heuristics that provides the best performance with respect to the three measures is identified. The heuristics include combinations of the shortest completion time and due date-based rules. Finally, a multi-criteria decision-making analysis is performed to determine the conditions specified by the weight given to each measure, with which one heuristic is preferred over the others.

Efficient I/O Performance-Focused Scheduling in High-Performance Computing

High-performance computing (HPC) systems are becoming increasingly important as contemporary exascale applications with demand extensive computational and data processing capability. To optimize these systems, efficient scheduling of HPC applications is important. In particular, because I/O is a shared resource among applications and is becoming more important due to the emergence of big data, it is possible to improve performance by considering the architecture of HPC systems and scheduling jobs based on I/O resource requirements. In this paper, we propose a scheduling scheme that prioritizes HPC applications based on their I/O requirements. To accomplish this, our scheme analyzes the IOPS of scheduled applications by examining their execution history. Then, it schedules the applications at pre-configured intervals based on their expected IOPS to maximize the available IOPS across the entire system. Compared to the existing first-come first-served (FCFS) algorithm, experimental results using real-world HPC log data show that our scheme reduces total execution time by 305 h and decreases costs by USD 53 when scheduling 10,000 jobs utilizing public cloud resources.

Intricate Supply Chain Demand Forecasting Based on Graph Convolution Network

Globalization has contributed to the increasing complexity of supply chain structures. In this regard, precise demand forecasting for the intricate supply chain holds paramount importance in effective supply chain management. Traditional statistical models and deep learning methods often face challenges in efficiently discerning correlations within a myriad of interconnected demands. To tackle this issue, this paper proposes an intricate supply chain demand forecasting method based on graph convolution networks adept at handling non-Euclidean data. First, the companies within the supply chain are treated as nodes in the graph structure, and the relationships between them are treated as edges, with demand data serving as the features of these edges. Then, a graph convolutional network is constructed to aggregate node and edge information. Through a multi-layer network, the relationships among nodes, edges, and historical demand are established to facilitate the prediction of supply chain demands. In this process, the graph convolutional network incorporates supply chain connectivity information into demand time series analysis. This integration of surface-level topological features and deeper latent correlation attributes across the supply chain’s nodes refines the demand forecasting precision across the entire supply chain. The validation experiment in this paper is grounded in sales data of a singular product from multiple sales nodes of an electronics company. The results demonstrate that the proposed method surpasses four other traditional demand forecasting algorithms significantly in terms of accuracy, providing substantial evidence for the superior performance of graph networks in the analysis of intricate supply chain relationships.

A Genetic algorithm approach for improving the probabilistic inventory model with the continuous review: A practical application in the department of pharmacy

This study investigates to shed light on artificial intelligence techniques, specifically the genetic algorithm, to improving traditional solutions, as well as finding the best policy for the problem of probabilistic inventory with continuous review by finding the optimal reorder point and the optimal economic quantity to avoid the risk of stock out (shortage), reduce total costs, and reach the optimal solution with little time and effort. The research was conducted in the stores of the department of pharmacy of the Ninawa health department for the period from January 1, 2021, to January 1, 2023, on a sample of three drugs (most demanded drugs). An analysis of the demand data for the study sample was conducted using the statistical program (SPSS Statistics Version 22) to determine the type of inventory. it was found that the type of inventory is probabilistic and the demand data follows a normal distribution. Based on this foundation, the mathematical model for the study problem was built. Given the complexity of the steps and iterations of the traditional solution, the solution steps were applied using the R programming language to reach the model's solution. Additionally, the solution steps were programmed for the genetic algorithm and applied using the R programming language. The results of the genetic algorithm showed an improvement in the traditional solution results by reducing total costs. Based on the results of the genetic algorithm, the economic order quantity, safety stock, reorder period, and safety period were found. Paper type: Research Paper

Techno-economic analysis of hybrid renewable energy systems for cost reduction and reliability improvement using dwarf mongoose optimization algorithm

The global energy crisis, particularly in isolated and remote regions, has increased interest in renewable energy sources (RESs) to meet growing energy demands. Integrating RESs with energy storage systems offers a promising solution to mitigate fluctuations and intermittency, but concerns about cost and reliability remain. This study explores the optimal design of various microgrid configurations, combining photovoltaic (PV), wind turbine (WT), battery energy storage system (BESS), and diesel generator (DG) systems for Najran city, Saudi Arabia, via real-world meteorological and load demand data. The Dwarf Mongoose Optimization Algorithm (DMOA), alongside the salp swarm algorithm (SSA) and whale optimization algorithm (WOA), was applied to minimize the levelized cost of energy (LCOE) while improving system reliability. The results demonstrate that the PV/BESS configuration, although cost-effective with an LCOE of 0.038 USD/kWh, fail to meet reliability constraints with a loss of power supply probability (LPSP) of 0.679. In contrast, the PV, WT, BESS, and DG configurations achieved an LPSP of 1.9 × 10^--8% with an LCOE of 0.199 USD/kWh, offering a robust and reliable solution for the region's energy needs. This paper presents a novel application of the DMOA for optimizing hybrid renewable energy systems, demonstrating its effectiveness in achieving a balance between cost and reliability. This strategy provides a viable approach for sustainable energy planning in similar regions facing energy challenges.

A novel data-driven rolling horizon production planning approach for the plastic industry under the uncertainty of demand and recycling rate

Efficient production planning in the plastic industry requires integrating sustainability practices, particularly the use of recycled plastics. Recycling helps reduce environmental impacts and conserve resources by decreasing the reliance on virgin materials. To develop a responsive plan, dynamic approaches are beneficial to confront fluctuations in uncertain parameters such as recycling rate and market demand. The present study proposes a novel Data-driven Rolling Horizon Planning (DRHP) approach for a sustainable and dynamic production plan in the plastic industry. A dynamic Rolling Horizon (RH) planning framework is formulated as a multi-product, multi-period Mixed Integer Linear Programming (MILP) model for the problem. This model aims to minimize total production cost while taking into account the system’s constraints and sustainability considerations. To deal with uncertainty, the RH-based MILP model is coupled with a rolling Long Short-Term Memory (LSTM) model. The rolling LSTM model leverages historical data of market demand and recycling rate to predict their future values and consequently improve the responsiveness of the production plan. The outperformance of the proposed DRHP approach is demonstrated through an extensive comparison with a robust static counterpart in terms of total production cost and sustainability performance. Results indicate significant reductions, up to 80%, in production costs using the proposed DRHP approach. Furthermore, the effect of rolling duration is investigated concerning production cost, backlog, late-order, and inventory level. Findings highlight the potential of the proposed DRHP approach to mitigate inventory- and late-order-related challenges.

Methods for projecting health workforce planning in Ireland: a collaborative approach

Abstract Background Health workforce planning (HWFP) is the process of analysing, forecasting, and planning workforce supply and demand, assessing gaps, and determining target workforce to ensure that health systems are prepared for future demand. HWFP elicits recommendations that are ideally utilised by the health system to support medical workforce over a defined time. Methods The objective of this project is to inform intake for medical and surgical specialties in Ireland to reach the necessary consultant workforce by 2038. A literature search was conducted to help inform the methodology for this project. The most common approach was a needs-based model, and additional reviews discussed the integration of medical engagement into the modelling approach. Results This project will employ a collaborative approach, combining mathematical modelling and a combination of quantitative and qualitative inputs, including the following phases: 1. The research phase will estimate baseline supply and demand data across medical and surgical specialties in Ireland using statistical models; 2. The presentation phase will include input from clinical leads to allow for more accurate assumptions in the baseline of each model; 3. Review of projections will include feedback from the clinical leads to help clarify and improve the initial model; 4. Presentation of the revised projection will include further collaboration with clinical leads to finalise the model. Conclusions The robust methodology for this HWFP project will ideally inform medical and surgical specialities in Ireland through 2038. Specific and extensive predictions of health workforce supply and demand will enhance strategic preparation, which is a key priority for the Health Service Executive in Ireland. Key messages • A collaborative approach to HWFP is essential for health service preparation and improvement. • The project will be finalised by the end of 2024, and the methodology for conducting similar HWFP projects can be adapted by other EU states.

Comparison between the models of potential storage and simulation of the demand for electricity transformer (400/11) in Diyala General Company

Storage models are usually used to determine the optimal order quantity and reorder point by obtaining the lowest total cost of storage to balance the amount of production and the amount of demand for materials, but sometimes these models can not be applied, so the use of simulation is resorted to where accurate and flexible results can be relied upon, to achieve the goal of the study, which is to improve the company's storage system. This research dealt with two cases of the probabilistic storage model, where the first model was applied to the monthly demand data on one of the products of Diyala General Company, which is the distribution transformer (400|11) for the years 2019,2020,2021. The second is the probabilistic storage model by simulating demand and comparing the two models, whichever is better to reach optimal. Where the distribution of demand was found to find the average and standard deviation and apply the model to obtain the indicators of the models. Using the MATLAB program, the statistical analysis of the data was carried out and the indicators of the model were obtained.

Risk Assessment of Urban Water and Energy Supply Using Copula Function: A Water–Energy Nexus Approach in an Arid City

Planning for the future of water and energy supply systems in urban areas requires a thorough assessment of associated risks. In this study, monthly water and energy demand data from 2011 to 2022 in an arid city was used to predict the corresponding demands from 2023 to 2032 using the seasonal auto-regressive integrated moving average (SARIMA) method. The aim is to estimate future water and energy supply risks both individually and jointly, using cumulative distribution functions (CDFs) derived from historical data. The main focus is to calculate the combined risk of water and energy, referred to as the water–energy nexus (WEN) risk. Based on the interdependent relationship between water and energy, the Copula function was utilized to model the bivariate distribution between these two variables. Pearson correlation analysis indicated a strong correlation between water and energy supplies. Among the distributions fitted to the data, the log-normal and gamma distributions were the best fit for water supply and energy supply systems, respectively, with the lowest Akaike information criterion (AIC) values. The Gumbel Copula, with a parameter of 1.66, was identified as the most suitable for modeling the joint distribution, yielding the lowest AIC value. The results indicate that the risks associated with energy supply, water supply, and their joint dependency could exceed 0.8% in the future, highlighting a potentially critical situation for the city. The trend analysis revealed that forecasted water and energy demands and their corresponding risks and the WEN risk are expected to have a significant upward trend in the future. Finally, local authorities need to explore alternative sources to supply water and energy in the future to address the ever-growing water and energy demands.