Search

The joint optimization problem of multiresource capacity planning and multitype patient scheduling under uncertain demands and random capacity consumption poses a significant computational challenge. The common practice in solving this problem is to first identify capacity levels and then determine patient scheduling decisions separately, which typically leads to suboptimal decisions that often result in ineffective outcomes of care. In order to overcome these inefficiencies, in this paper, we propose a novel two-stage stochastic optimization model that integrates these two decisions, which can lower costs by exploring the coupling relationship between patient scheduling and capacity configuration. The patient scheduling problem is modeled as a Markov decision process. We first analyze the properties for the multitype patient case under specific assumptions and then establish structural properties of the optimal scheduling policy for the one-type patient case. Based on these findings, we propose optimal solution algorithms to solve the joint optimization problem for this special case. Because it is intractable to solve the original two-stage problem for a general multitype system with large state space, we propose a heuristic policy and a two-stage stochastic mixed-integer programming model solved by the Benders decomposition algorithm, which is further improved by combining an approximate linear program and the look-ahead strategy. To illustrate the efficiency of our approaches and draw managerial insights, we apply our solutions to a data set from the day surgery center of a large public hospital in Shanghai, China. The results show that the joint optimization of capacity planning and patient scheduling could significantly improve the performance. Furthermore, our model can be applied to a rolling-horizon framework to optimize dynamic patient scheduling decisions. Through extensive numerical analyses, we demonstrate that our approaches yield good performances, as measured by the gap against an upper bound, and that these approaches outperform several benchmark policies. Summary of Contribution: First, this paper investigates the joint optimization problem of multiresource capacity planning and multitype patient scheduling under uncertain demands and random capacity consumption, which poses a significant computational challenge. It belongs to the scope of computing and operations research. Second, this paper formulates a mathematical model, establishes optimality properties, proposes solution algorithms, and performs extensive numerical experiments using real-world data. This work includes aspects of dynamic stochastic control, computing algorithms, and experiments. Moreover, this paper is motivated by a practical problem (joint management of capacity planning and patient scheduling in the day surgery center) in our cooperative hospital, which is also key to numerous other applications, for example, the make-to-order manufacturing systems and computing facility systems. By using the optimality properties, solution algorithms, and management insights derived in this paper, the practitioners can be equipped with a decision support tool for efficient and effective operation decisions.

When a fixed number of support vector machines (SVMs) are taken as the base learners, an attempt to diversify them should be encouraged to achieve a satisfactory ensemble. In this article, by means of a negative agreement learning (NAL) strategy, a new SVM-based ensemble framework is proposed to simultaneously enhance the diversity of SVMs in the ensemble and suppress the training error of the ensemble. The proposed ensemble framework is theoretically derived to have distinctive merits: 1) the ensemble and each of its individual SVM base learner are trained in a joint manner rather than in an independent manner and 2) the NAL strategy facilitates the formulation of the ensemble of SVMs as one single SVM; thus, abundant advances in the training of SVM can be conveniently applied to the proposed ensemble learning of SVMs and there is no need to design special optimization techniques for the involved ensemble learning. Extensive experimental studies demonstrate the effectiveness of the proposed ensemble framework of SVMs.

In this study, we propose a new appointment window scheduling (AWS) approach of informing customers of an admission window (AW) rather than the traditional appointment time. We provide a formal des...

Reverse logistics networks reintroduce end-of-life products to remanufacturing, which is significant for sustainable development and environmental protection. In this paper, we investigate the curr...

Epileptic EEG signal recognition is an important method for epilepsy detection. In essence, epileptic EEG signal recognition is a typical imbalanced classification task. However, traditional machine learning methods used for imbalanced epileptic EEG signal recognition face many challenges: (1) traditional machine learning methods often ignore the imbalance of epileptic EEG signals, which leads to misclassification of positive samples and may cause serious consequences and (2) the existing imbalanced classification methods ignore the interrelationship between samples, resulting in poor classification performance. To overcome these challenges, a graph-based extreme learning machine method (G-ELM) is proposed for imbalanced epileptic EEG signal recognition. The proposed method uses graph theory to construct a relationship graph of samples according to data distribution. Then, a model combining the relationship graph and ELM is constructed; it inherits the rapid learning and good generalization capabilities of ELM and improves the classification performance. Experiments on a real imbalanced epileptic EEG dataset demonstrated the effectiveness and applicability of the proposed method.

Manufacturing servitization has become a major trend in industry that is implemented by so called integrated production and service systems (IPSSs) to offer not only products but also their associated services. In this study, we explore the capacity allocation policy for an IPSS that consists of a manufacturing facility and a service center. The IPSS serves two classes of customers, each demanding a specific product produced at the manufacturing facility and, subsequently, the associated service offered by the service center. We formulate this problem of resource allocation as a Markov decision process. Our analysis suggests that it is optimal to serve the customer with the larger cost saving rate and slower service rate, where the cost saving rate equals service rate times the sum of holding cost and waiting cost. Then we explore production control policies coupled with optimal service ones. We demonstrate that, with light service traffic, the system under the optimal production policy becomes similar to a make‐to‐stock system. Such a system can effectively hedge the system uncertainty. Therefore, the optimal production policy for the IPSS is also a hedging point policy. Under this policy, switching and idling curves split the state space into three regions: one without any production, the two others for production of two type products, respectively. We also discuss the optimal control policies of the two extensions of the IPSS (one with multiple types of product and customers, and the other with a single product and two classes of customers) and find that the characters of the optimal production and service policy still hold. Finally, we develop three heuristic integrated scheduling policies. Numerical experiments show that one is very effective.

This paper investigates the station-keeping control of autonomous and remotely-operated vehicles (ARVs) for free-floating manipulation under model uncertainties and external disturbances. A modified adaptive generalized super-twisting algorithm (AGSTA) enhanced by adaptive tracking differentiator (ATD) and reduced-order extended state observer (RESO) is proposed. The ATD is used to obtain the smooth reference signal and its derivative. The RESO is used to estimate and compensate for the model uncertainties and external disturbances in real-time, which enhances the robustness of the controller. The modified AGSTA ensures the fast convergence of the system states and maintains them in a predefined neighborhood of origin without overestimating control gains. Besides, the proposed new variable gain strategy completely avoids the control gains vibrating near the set minimum value. Thanks to the RESO, the proposed controller is model-free and can be easily implemented in practice. The stability of the closed-loop system is analyzed based on Lyapunov’s direct method in the time domain. Finally, the proposed control scheme is applied to the station-keeping control of Haidou-1 ARV, and the simulation results confirm the superiority of the proposed control scheme over the original AGSTA.

With the aim of supporting future traffic needs, an account of how to reconstruct an existing cyclic timetable by inserting additional train services will be given in this paper. The Timetable-based Extra Train Services Inserting (TETSI) problem is regarded as an integration of railway scheduling and rescheduling problem. The TETSI problem therefore is considered involving many constraints, such as flexible running times, dwell times, headway and time windows. Characterized based on an event-activity graph, a general Mixed Integer Program model for this problem is formulated. In addition, several extensions to the general model are further proposed. The real-world constraints that concerning the acceleration and deceleration times, priority for overtaking, allowed adjustments, periodic structure and frequency of services are incorporated into the general model. From numerical investigations using data from Shanghai-Hangzhou High-Speed Railway in China, the proposed framework and associated techniques are tested and shown to be effective.

This article addresses the periodic home health care (HHC) server assignment problem in HHC companies. This problem is a variant of the periodic vehicle routing problem with specific constraints for customers and servers (e.g., customers’ requirements for multiple medical skills and continuous care offered by a given number of servers and servers’ requirements for workload balance). Solving large-scale problems from practical applications of HHC companies within an acceptable computation time is challenging. Consequently, we develop an efficient region-partition-based algorithm to solve these large-scale problems. First, the algorithms assign customers and servers to many independent regions. Second, four different tabu search (TS) algorithms are designed to solve the optimization problem of each region. Finally, the algorithms iteratively adjust the assignment of customers and servers to regions and solve the problems of each region. The performances of different TS algorithms are discussed. The effectiveness of the proposed region-partition-based algorithm for large-scale problems is validated. Note to Practitioners —This article is motivated by our collaboration with an HHC company in Shanghai, China. The company delivers medical, paramedical, and social services to customers in their homes to help them improve their clinical and psychological conditions without hospital stays. In practice, customers require long-term and periodic service (e.g., weekly), which may last many months or even years. HHC companies have to assign a combination of service days and at least one server to each customer. Due to specific constraints in practical operations and the large scale of the practical problems (e.g., 1000–2000 customers) of the HHC in this article, the server assignment problem is challenging to the HHC company. The algorithm presented in this article takes into account the special constraints of the HHC and is able to solve the large-scale problem in a reasonable time span. The algorithm can be used to help decision-makers assign combinations of service days and servers to customers with the objective of balancing the workloads among servers.

Hospital beds are one of the most critical resources in healthcare institutions. In practice, beds are usually allocated to different departments in advance to accommodate different kinds of patien...

Cost-effectiveness analysis is an important topic in public health, which can provide valuable information for medical decisions. Several modeling methods are available for conducting cost-effectiveness analysis. However, it is difficult when the data is incomplete. To solve this problem, a Markov model is proposed to model patients’ health states transition, and two hybrid metaheuristics are proposed to estimate the transition probabilities. Based on the estimated transition probabilities, cost-effectiveness analysis is conducted to compare different medical interventions. Numerical experiments and case study validate the effectiveness and practicability of the proposed method. The case study gives the physicians effective instructions by comparing two different immunosuppressants after renal transplantation.

When handling special multi-view scenarios where data from each view keep the same features, we may perhaps encounter two serious challenges: (1) samples from different views of the same class are less similar than those from the same view but different class, which sometimes happen in local way in both training and/or testing phases; (2) training an explicit prediction model becomes unreliable and even infeasible for test samples in multi-view scenarios. In this study, we prefer the philosophy of the k nearest neighbor method (KNN) to circumvent the second challenge. Without an explicit prediction model trained directly from the above multi-view data, a new multi-view local linear k nearest neighbor method (MV-LLKNN) is then developed to circumvent the two challenges so as to predict the label of each test sample. MV-LLKNN has its two reliable assumptions. One is the theoretically and experimentally provable assumption that any test sample can be well approximated by a linear combination of its neighbors in the multi-view training dataset. The other assumes that these neighbors should demonstrate their clustering property according to certain commonality-based similarity measure between the multi-view test sample and these multi-view neighbors so as to avoid the first challenge. MV-LLKNN can realize its effective prediction for a test multi-view sample by cheaply using both on-hand fast iterative shrinkage thresholding algorithm (FISTA) and KNN. Our theoretical analysis and experimental results about real multi-view face datasets indicate the effectiveness of MV-LLKNN.

This paper considers the problem of minimising makespan on a single batch processing machine with flexible periodic preventive maintenance. This problem combines two sub-problems, scheduling on a b...

This paper introduces a special vehicle routing problem, i.e. the cumulative capacitated vehicle routing problem with time-window constraints (Cum-CVRPTW). The problem can be defined as designing least-cost delivery routes from a depot to a set of geographically-scattered customers, subject to the constraint that each customer has to be served within a time window; accordingly, the objective costs are computed as the sum of arrival times at all the customers. The Cum-CVRPTW finds practical utility in many situations, e.g. the provision of humanitarian aids in the context of natural disasters. The Cum-CVRPTW can be viewed as a combination of two NP-hard problems, i.e. the vehicle routing problem with time windows and the cumulative vehicle routing problem. To effectively address this problem, an effective algorithm is designed, which is based on the frameworks of Large Neighborhood Search Algorithm and hybridizes with Genetic Algorithm. The proposed algorithm adopts a constraint-relaxation scheme to extend the search space, enabling the iterative exploration of both the feasible and infeasible neighborhood solutions of an incumbent solution. Furthermore, some speed-up techniques are designed to reduce the computational complexity. To elucidate its effectiveness, the proposed algorithm is examined on the benchmark instances from the literature. The resultant numerical findings show that the algorithm is able to improve and obtain some best-known solutions found by existing state-of-the-art methods.

In the real manufacturing system, rescheduling is inevitable because of rush orders. To improve the rush order inserting problem of rescheduling, this paper focuses on the just-in-time job-shop rescheduling problem (JIT JSRP), in which each job has its own due date and any earliness /tardiness leads to the penalty. A mixed integer programming model is established to minimize the weighted penalty cost of earliness/tardiness and the starting time deviations. The paper develops a hybrid tabu-variable neighborhood search (HTVNS) algorithm to solve the problem. Moreover, the adaptive shake operator selection algorithm and two improved N5 neighborhood structures are introduced to improve the efficiency of the algorithm. In numerical experiments, the improved algorithm is testified using 36 cases with different scales and arrival times of rush orders, and compared with classical meta-heuristic algorithms. The computational results show the effectiveness of the proposed improved algorithm.

Cellular reconfigurable manufacturing systems (CRMS) are designed to deal with dynamic demand. This paper tries to optimize cell formation decision in CRMSs with alternative routing. A mixed integer nonlinear programming model is developed to simultaneously optimize cell formation, equipment configuration, and product mix plan. The objective is to maximize profit. The linearization methods are proposed and then the model is solved by Gurobi. Numerical experiment and sensitive analysis are conducted to show the efficiency and effectiveness of the proposed approach.

In the situation of intelligent manufacturing, the demand for individual customization increases. Many enterprises adopt make-to-order production. With the increasing trend of globalization, numerous enterprises are multinational, which have multi-location plants and distributors. The customized orders are split into multiple different jobs and then produced by different plants. Under the requirements of low-cost production and rapid response to demand, the coordination between production and transportation is becoming increasingly important. Considering the features of order splitting production, heterogeneous vehicle transportation, and collaborative delivery in make-to-order production, this paper studies the integrated scheduling of production and transportation of multiple splittable orders in networked plants and distributors with different geographical locations. A mixed-integer linear programming model is developed for deciding the production plant, start production date, the transport vehicles, and departure date of each job from its assigned plant with the objective of minimizing total cost including manufacturing cost, inventory cost, transportation cost, and backorder cost. The commercial solver Gurobi is used to solve this model. Numerical experiments based on the real data of a customized furniture enterprise in China are performed to show the efficiency of the proposed approach and demonstrate that our integrated scheduling approach is more effective than sequential scheduling.

In the context of manufacturing servitization, manufacturing enterprises offer not only products but also associated services according to customer demands, which is the so-called Product Service System (PSS). In PSS configuration, enterprises need to jointly determine the selected modules with corresponding prices to improve sales profit and customer satisfaction. Better customer satisfaction can be achieved by lower price and larger module utility. Apart from most previous research, we consider not only the single module utility but also the pairing utility, which measures the additional value jointly created by a product module and a service module to illustrate the synergy effect of “product + service”. Furthermore, we take the uncertainty of service utility and customer behavior into account. The former is due to the heterogeneity and intangibility nature of service, and the latter is described by price sensitivity that varies among different customers. To realize the above two contributions, we first establish a multi-objective stochastic programming model to maximize expected customer satisfaction as well as sales profit. For a better trade-off between solution quality and computation efficiency, we develop an improved algorithm based on epsilon-constraint and sample average approximation, including the feasibility examination and accelerating strategy. Finally, numerical experiments are conducted to demonstrate our approach and validate the contributions.

In narrow overlap welding, serious defects in the weld will lead to band breakage accident, and the whole hot dip galvanizing unit will be shut down. Laser vision inspection hardware is used to collect real-time image of weld surface, and image defect recognition and evaluation system is developed to real-time detect quality. Firstly, region division is implemented. In view of the characteristics of gray image such as large information, low contrast and blurred edge, an improved image segmentation algorithm is proposed by using image convolution to enhance edge features and combining with integral image, which can quickly and accurately extract the weld edge and divide the region, and the processing time can meet the real-time requirements. Then comparing the effect of traditional method and convolution neural network in identifying defects, VGG16 is used to identify weld defects. In order to ensure real-time performance, a two-stage weld defect recognition is proposed. First, the large defective area is identified, and then, the defect is accurately identified in the area. This method can quickly extract defect areas and complete weld defect classification. Experiments show that the accuracy can reach 97% and average running time takes 3.2 s, meeting the online detection requirements.

As a new manufacturing paradigm, Reconfigurable Manufacturing System (RMS) has a modular structure, which can greatly improve the responsiveness and flexibility by adding or removing the modules of the reconfigurable machine tools (RMTs). However, the uncertain demand makes it difficult and challenging to configure the RMS. To deal with this problem, we propose a multi-period stochastic programming model to optimize the configuration of RMS in each period by considering the uncertain demand. The objective is to maximize total costs. Production planning related decisions are considered to better handle the unpredictable demand fluctuation. Numerical experiments compare the solutions by solving the proposed stochastic programming model and those by solving the deterministic linear model without considering production planning decisions. The results show that our model can provide a better solution in terms of the configuration and reconfiguration cost and total costs under the consideration of uncertain demand.

Precise and reasonable evaluation of the multi-attribute value of medical system is the basis for hospitals to implement total quality management. Excellent medical system is necessary as a part of modern urban governance. However, most of medical value evaluation work relies on scale and artificial scoring at present, lacking in objectivity. Therefore, a scientific and comprehensive medical value evaluation system is needed urgently to give full play to the guiding role of value evaluation and promote the improvement of the medical service system. In this study, DEMATEL and information entropy are used to quantify the degree of mutual influence between system indicators and the differences in medical market performance respectively, so as to obtain the objective index weight. Hypertension has the highest incidence in the whole chronic disease system, which seriously affects people's daily life. Based on the existing hypertension diagnosis and treatment index system, a comprehensive and objective evaluation model is established to evaluate the hypertension diagnosis and treatment behaviors of different medical institutions, which achieves good result. This method has effectively improved the relative deficiency of one-sided subjective evaluation and has a great guiding significance for the comparison of treatment in departments and the economical use of medical resources.

This study presents a methodology for estimating passenger’s spatio-temporal trajectory with personalization and timeliness by using incomplete Wi-Fi probe data in urban rail transit network. Unlike the automatic fare collection data that only records passenger’s entries and exits, the Wi-Fi probe data can capture more detailed passenger movements, such as riding a train or waiting on a platform. However, the estimation of spatio-temporal trajectories remains as a challenging task because a few unfavorable situations could result into deficient data. To address this problem, we first describe the Wi-Fi probe data and summarize their common defects. Then, the n-gram method is developed to infer missing spatio-temporal location information. Next, an estimation algorithm is designed to generate feasible spatio-temporal trajectories for each individual passenger by integrating multiple data sources, i.e., urban rail transit network topology, Wi-Fi probe data, train schedules, etc. This proposed method is tested on both simulated data in blind experiments and real-world data from a complex urban rail transit network. The results of case study show that 93% of passengers’ unique physical routes can be estimated. Then, for 80% of passengers, the number of feasible spatio-temporal trajectories can be reduced to one or two. Potential applications of the trajectory estimation approach are also identified., Highlights • This study extends personalization and timeliness into the spatio-temporal trajectory estimation by integrating structurally heterogeneous data sources. • A new algorithm of estimating individual passenger’s spatio-temporal trajectory that describes detailed information both on-board and at stations during the trip is developed. • Real-world analyses are conducted for the urban rail transit network with the largest route length in China. • This is the first attempt to find out all the routes and trains that passengers actually choose using Wi-Fi probe data at the network level.

This paper aims to devise efficient combination chemotherapy schedules that determine the dosages of drugs administered to cancer patients with drug resistance that, as a Gordian knot to cancer chemotherapy, may weaken the efficacy of chemotherapy. To characterize cell growth, we use the existing cell cycle-specific model, in which the mechanism of acquired drug resistance is incorporated. Subsequently, the determination of the optimal chemotherapy schedule for the patients is formulated as a nonlinear optimization problem, with the objective of minimizing not only the quantity of tumor cells but also the posttreatment chemotherapy-induced toxicity. To overcome the difficulty in finding a satisfactory solution to the problem due to its nonlinear nature, we develop a memetic algorithm (MA) with an advanced local search strategy. The efficiency of the proposed MA is validated by comparison with other state-of-the-art methods. In addition, we compare the best found solution to the problem in the presence of drug resistance with that in the absence of drug resistance. The resultant findings reveal that drug resistance is a crucial factor in the determination of chemotherapy schedules. Note to Practitioners —The efficacy of chemotherapy is limited by drug toxicity and is weakened by drug resistance. This paper deals with the problem of optimization of chemotherapy schedules with mathematical modeling and heuristic algorithms and is the first to consider the drug specificity, drug combination, and drug resistance simultaneously. The proposed algorithm is flexible. It can be applied to other optimization problems with proper adaptations. The experimental results suggest that the best found solution to the problem in the absence of drug resistance may, however, not remain as the best found solution, or even become an infeasible one, in the presence of drug resistance. Existing clinical regimens may be improved with the assistance of our method that aims to optimize the chemotherapy schedule in the presence of drug resistance. The parameters of our model should be refined with experimental and clinical data to confer more benefits on real treatments in clinical practice in the future.

In the case of an over-crowded urban rail transit (URT) line, a large number of passengers may be left stranded and daily timetable may become infeasible. This paper proposes a coordinated optimization scheme for a URT line, which combines both the coordinated passenger inflow control with train rescheduling strategies. With the aim of minimizing the penalty value of passengers being stranded along the whole line, the coordinated passenger inflow control helps relieve demand pressure and ensure safety at over-crowded URT stations while the train rescheduling of skip-stopping helps to balance the utilization of train capacity. A novel Q-learning based approach to this combination optimization problem is developed. Simulation experiments are carried out on a real-world URT line in Shanghai. Basic principles of Q-learning are presented, which consist of the environment and its states, learning agents and their respective actions, and rewards. The results show that the coordinated optimization scheme solved by the Q-learning approach is effective in relieving the passenger congestion on the URT line. The Q-learning approach can offer accurate scheme to deal with the problem of passenger congestion and train operation on a URT line.

This article studies the optimal allocation of inpatient rooms for multiple types of patients in public hospitals and the patient scheduling problem with planned acceptance ratios (ARs). For public hospitals, it is important to allocate limited resources to multiple types of patients and manage patient access for maximizing hospital revenue and upholding service equity. The problem is formulated as two-stage models. Considering uncertainties in patients’ arrival and length of stay, we first propose a nonlinear stochastic programming (NSP) model for inpatient room allocation with the objective of maximizing revenue under the constraints of maintaining equity. To solve this problem, we transform the complex NSP model into a deterministic mixed-integer linear programming model, which is solved by CPLEX, by reformulating the chance constraints as knapsack constraints based on a linearization technique and a simulation model. Given the allocated capacity and planned AR, we further propose a two-stage stochastic mixed-integer program combined with a goal program model to optimize patient scheduling. To solve the model, a Benders decomposition based on the sample average approximation approach is proposed. The real data-based experimental results demonstrate the applicability and effectiveness of our models and approaches. The impacts of some parameters on the objective and decisions are also explored. A simulation procedure is developed to compare the performances of different patient scheduling methods, from which the results show that our proposed approach outperforms a benchmark policy. Note to Practitioners —Inpatient rooms are critical resources for hospitals. Against the background of aging populations and environmental problems, the twofold predicament—involving escalating healthcare demands and insufficient room-based resources—has led to the necessity for hospitals to operate more effectively and efficiently. Capacity management and patient scheduling are thus the two most important operations for hospital management. Because of the self-financing feature of hospitals and the quasi-public nature of medical services, it is important for public hospitals to judiciously allocate limited room capacities to multiple types of patients for balancing revenue and equity and schedule the arrival demands dynamically according to the planned capacity and acceptance ratio. In this article, we propose mathematical models and solution approaches for these two-stage problems, and their applicability and effectiveness are demonstrated by experiments based on real data. The managerial insights suggest that hospitals should improve the service equity gradually according to their financial situation because of the increasing marginal cost and should apply the scientific approaches and techniques, rather than by their experiences, to aid their management for better performance. By using approaches proposed in this article, hospital managers can be equipped with a decision support tool for effective capacity allocation and patient scheduling decisions. The parameters of our models could be tuned based on the preferences of different hospitals. Furthermore, these approaches can be applied to other settings with similar problems, such as government budget allocation considering both utility and equity, and service system management with waiting time requirement.

As service is becoming the strategic preference in the manufacturing industry, more and more manufacturing companies provide customers with total solutions or product service system by integrating services into their core product offerings, which is usually implemented in so-called hybrid supply chain. The hybrid supply chain is formed as a chain by multi-echelon manufacturers and service providers who work together to offer product service system to customers. In multi-echelon hybrid supply chain, each echelon customers’ response to variable service levels and interaction between inventory and service capacity strategy have a great impact on supply chain instability, which represents an important concern in supply chain research. To tackle this problem, we establish a system dynamics model, by taking the multi-echelon hybrid supply chain for a logistics equipment company in China as an example. Firstly, we analyzed the oscillation characteristics of service flow and product flow. Then, we proposed the performance metrics of bullwhip effect in multi-echelon hybrid supply chain. Finally, based on the model, we found that the bullwhip effect of multi-echelon hybrid supply chain could be smoothened by incorporating forecast smooth factor, control percentage coefficient and order lead time.

Real-time anomaly detection at metro stations is a very important task with considerable implications for massive passenger flow organization and train timetable rescheduling. State-of-the-...

This study explores two nonparametric machine learning methods, namely support vector regression (SVR) and artificial neural networks (ANN), for understanding and predicting high-speed rail (HSR) travelers’ choices of ticket purchase timings, train types, and travel classes, using ticket sales data. In the train choice literature, discrete choice analysis is the predominant approach and many variants of logit models have been developed. Alternatively, emerging travel choice studies adopt non-utility-based methods, especially nonparametric machine learning methods including SVR and ANN, because (1) those methods do not rely on assumptions on the relations between choices and explanatory variables or any prior knowledge of the underlying relations; (2) they have superb capabilities of iteratively identifying patterns and extracting rules from data. This paper thus contributes to the HSR train choice literature by applying and comparing SVR and ANN with a real-world case study of the Shanghai-Beijing HSR market in China. A new normalized metric capturing both the load factor and the booking lead time is proposed as the target variable and several train service attributes, such as day of week, departure time, travel time, fare, are identified as input variables. Computational results demonstrate that both SVR and ANN can predict the train choice behavior with high accuracy, outperforming the linear regression approach. Potential applications of this study, such as rail pricing reform, have also been identified.

Hospital wards are critical resources because of limited space and large construction investment and operating costs. They are shared among different types of patients with different access time targets determined by their diseases and payments. In countries like China, it is important for public hospitals to simultaneously maximize hospital revenue as well as equity among different types of patients when allocating these limited capacities. Consequently, hospital managers are under high pressure to consider different types of patients’ access time targets and allocate them wards without decreasing revenues. To address this problem, a multi-objective stochastic programming (MSP) model is proposed with the objective to maximize both revenue and equity. Random patient arrivals and lengths of stay make it difficult to analytically describe both revenue and equity objectives in the MSP model. To cope with this problem, a data-driven discrete-event simulation model is proposed to find the relationship between model objectives regarding system performance and decisions regarding capacity allocation and patient admission. Then, based on the simulation results, we propose a linearization approach to transform the complex multi-objective stochastic programming model to a multi-objective integer linear programming (MILP) model, and an adaptive improved e-constraint algorithm and a multi-objective genetic algorithm combined with neighborhood search algorithm are proposed to solve the MILP problem. Based on the real data collected from a large public hospital in Shanghai, extensive numerical experiments are performed to demonstrate the efficiency of the model and solution approaches.

This paper addresses the solution methods for the stochastic home-caregiver scheduling and routing problem which arises in many service industries such as home care and home health care. In the problem, the caregiver’s travel times and service times for customers are stochastic. A chance constraint is introduced to ensure the on-time service probability for the customers. Such stochastic travelling and service time and the chance constraint further complicate the problem. In this paper, a route-based mathematical model is introduced. A branch-and-price (B&P) algorithm and a discrete approximation method are combined to solve the problem. Herein, effective label algorithms are designed to generate negative reduced cost routes. The efficiency of the algorithm are improved by employing three acceleration strategies. The experiments on test instances validate the performances of the proposed B&P algorithm and demonstrate the necessity of considering the stochasticity of travel times of home-caregiver and service times to the customers.

We introduce the load-dependent vehicle routing problem with time windows (LDVRPTW) in this paper. Transportation costs in this new problem, unlike those in the classical vehicle routing problem with time windows (VRPTW), are calculated based on not only the travel distances but also the vehicular loads on travel arcs. To solve this challenging NP-hard problem, we design a new constraint relaxation-based algorithm. In the proposed algorithm, a new constraint relaxation is introduced, i.e., some clients are not visited by a real vehicle and instead are entrusted to an additional virtual vehicle. Based on this relaxation, we present an effective execution scheme of local search procedures. The proposed algorithm is tested on benchmark instances of several special cases of the LDVRPTW, including the VRPTW. Numerical results for different variant problems demonstrate that the algorithm consistently yields impressive results: in particular, for one special variant, namely the fuel consumption rate considered vehicle routing problem (FCR-VRP), the algorithm improves the best-known solutions found by existing state-of-the-art methods.

Motivated by the practical requirements for timetable evaluation in China and the fact that massive train record data are collected but remain largely underused, this study presents a data analytics approach for train timetable performance measures using automatic train supervision data. A data preparation process with data cleaning and matching methods is developed for further analysis. The data analysis consists of three components: a waiting time assessment method that uses visual headway mismatching degree to estimate the effect on waiting time; a process time estimation method that introduces spatiotemporal distribution and statistical techniques to mine the data and identify practical characteristics of dwell time and running time; and an arrival punctuality examination method which checks on-time arrival performance. The proposed data analytics approach is demonstrated through a case study of Shanghai Metro. Major findings on timetable performance, involving three aspects of timetable parameters, are presented. The relevant data analytics framework and findings have operational and planning implications for urban rail transit authorities and operators with regard to evaluating timetable parameters and improving the service quality.

This paper addresses a daily caregiver scheduling and routing problem arising in home health care or home care service providers. The problem is quite challenging due to its uncertainties in terms ...

This paper provides a formal framework for optimal dynamic appointment scheduling with wait-dependent abandonment in order to reduce the blind wait (wait without queue information) and useless wait (wait but abandon). A discrete-time Markov decision process model is proposed to best balance server utility and experiences of customers waiting for admission. The properties of the optimal policy are investigated both theoretically and numerically. Efficient policies are implemented, the benefit from instituting the appointment system is checked numerically and managerial insights derived. In particular, appointment scheduling is shown to significantly reduce blind and useless wait without sacrificing the server utility when the system is overloaded with long service time and customers are more sensitive to blind wait than known wait (wait with a target). A revenue-oriented system tends to inform earlier appointment times and hence higher effective admission rate than a system taking into account customer experience of wait.

The problem with optimizing cancer chemotherapy has often been formulated as optimal control models, which can be intractable. Such difficulty is compounded by the facts that chemotherapeutic drugs with different mechanisms of action are often used together and that dose adjustments are often warranted according to therapeutic responses in clinical practice. Against this background, this paper addresses the problem of combination chemotherapy with dose adjustment. We first construct a mathematical model of the problem by introducing two cell-cycle phase-specific chemotherapeutic drugs into a mono-chemotherapy model. Next, we design a memetic algorithm (MA) allowing dose adjustments to solve the problem. Finally, we compare the proposed MA with existing algorithms, investigate the efficacies of different treatment strategies, and identify the characteristics of the problem related to the quality of the solutions.

In peak hours, when the limited transportation capacity of urban rail transit is not adequate enough to meet the travel demands, the density of the passengers waiting at the platform can exceed the critical density of the platform. Coordinated passenger inflow control strategy is required to adjust/meter the inflow volume and relieve some of the demand pressure at crowded metro stations so as to ensure both operational efficiency and safety at such stations for all passengers. However, such strategy is usually developed by the operation staff at each station based on their practical working experience. As such, the best strategy/decision cannot always be made and sometimes can even be highly undesirable due to their inability to account for the dynamic performance of all metro stations in the entire rail transit network. In this paper, a new reinforcement learning-based method is developed to optimize the inflow volume during a certain period of time at each station with the aim of minimizing the safety risks imposed on passengers at the metro stations. Basic principles and fundamental components of the reinforcement learning, as well as the reinforcement learning-based problem-specific algorithm are presented. The simulation experiment carried out on a real-world metro line in Shanghai is constructed to test the performance of the approach. Simulation results show that the reinforcement learning-based inflow volume control strategy is highly effective in minimizing the safety risks by reducing the frequency of passengers being stranded. Additionally, the strategy also helps to relieve the passenger congestion at certain stations.

Effective early diagnosis of autism can have a significant impact on its intervention and treatment. In this paper, an approach is proposed for comprehensively considering genetic factors and environmental factors to predict the severity of autism. According to the Childhood Autism Rating Scale (CARS), a sample set was collected from the autism clinic and a predictive model based on a stacked sparse autoencoder combined with a softmax classifier was constructed. We compared the proposed model with decision trees and support vector machines. Experiments show that the proposed model has a highest accuracy in predicting the severity of autism. Our method can help patients predict their condition and assist doctors in accurate diagnosis.

Current research on layout planning of grid seldom takes photovoltaic self-generating into consideration and rarely optimizes the substation and energy storage station (ESS) together. We propose an integrated optimization method on location and capacity planning of substation and ESS in photovoltaic self-generating grid in this paper. We first describe pattern of user power consumption and the photovoltaic self-generating power, before predicting the net load. Meanwhile, we derive time series of substation power distribution with the historical data. On this basis, we take the substation power distribution and capacity upper bound in consideration to obtain the substation capacity. And the time series of net load and substation power distribution together serves as the foundation to calculate the capacity of ESS. Then, we construct an integrated optimization model to maximize the profit of investment on substation and ESS. For simplification, we assume that there is only one ESS in the power grid, and the location is given. In the numerical experiment, the optimal location and capacity of substations, and the optimal capacity of ESS are obtained by CPLEX solver. Compared with a model without ESS, the proposed model is proved to improve the profit and the service scale of substations, which can enhance the performance of operation management under renewable energy environment.

In many countries, medical services are delivered through a multi-hospital network where a patient may have unlimited choices to different points of access to care. Due to various reasons, such a network may experience substantial workload imbalance. One way to address this challenge relies on government providing higher subsidy to incentivize patients to visit hospitals with low workload/utilization. In this research, we studied the problem of optimal government-to-patient subsidy differential (G2P-SD) policy design. We first formulated the problem with a nonlinear optimization model to minimize the total social cost (i.e., the cost of weighted wait time plus government subsidy spending) subject to the minimum workload requirement. Then we used a discrete choice model with real-world data to identify the significant influence of G2P-SD on patient hospital visit choice and numerically specified the rates of patient arrivals at a multi-hospital outpatient care network accordingly. We next developed a large-scale two-level queuing network to analyze the impact of G2P-SD on patient flows within the service network. We defined funding efficiency as a potential indicator to policy makers for effective budget allocation among various types of patients. Our study verified the effectiveness of modifying the G2P-SD policy, i.e., the total social cost is reduced by 55.99%. Furthermore, our study suggested the benefit of further tailoring the policy design with consideration of influential patient attributes, which leads to a further reduction in wait time at high-workload hospitals in our Shanghai-based case study.

This paper focuses on the scheduling problem of batch processing workstation (BPW) of β1→β2 type in semiconductor wafer fabrication (SWF). The constraints consist of incompatible families, limited waiting time, and re-entrant flow. BPW is a bottleneck in SWF. In order to ensure BPW that can process full-batch continuously, a new pull-based scheduling algorithm is proposed. According to collaborative strategy, the pull-based scheduling algorithm consists of three sub-algorithms. They are one-to-one forming batch sub-algorithm, two-to-one forming batch sub-algorithm, and pulling preceding jobs sub-algorithm. The experiments are implemented on a scheduling simulation platform of SWF. In the whole scheduling horizon, the three sub-algorithms are executed circularly. The results show that the proposed approach obtains better solutions.

With the urban development and enlargement, various regions such as residential zones and administrative districts now appear as parts of cities. People exhibit different mobility patterns in each region, which is closely relevant to region-wide functions. In this article, we propose a scheme to discover region-wide functions using large-scale Shanghai taxicab trajectories that capture enormous traces for more than 13,000 taxicabs over a period of about 3 years. We investigate these taxicab trajectories and conduct an extensive preliminary study. Then, we divide the city into disjointed regions using Voronoi decomposition. By incorporating people's pick-up and drop-off information, we refine the Voronoi partitioning results to identify region-wide functional areas. Finally, we study people's movement frequency on weekdays and weekends for every kind of urban functional regions. We also look into human mobility within or across the identified urban functional regions. Experimental results show that human movement is bounded with the function of urban regions, and more than 90% of people visit neighboring (less than 20km travel distance) functional regions with high probability.

When handling imbalanced datasets, the existing least learning machine ignores both the interrelationship between instances and the prior knowledge about the classes and hence tends to provide unfavorable accuracies across the classes. With the Laplacian matrix based loss function, Laplacian least learning machine (L2MM) is developed to address the infeasibility of the existing least learning machine for imbalanced data classification in this study. In order to find out the optimal number of hidden nodes in L2MM, by means of the derived update equations without any re-calculation of inverse matrix, its two incremental versions (i.e., with and without regularization) are invented for dynamically updating the hidden nodes in one-by-one way, thereby saving much training time. L2MM and its incremental versions inherit fast learning and good generalization capability of least learning machine. Experimental results on 19 benchmarking imbalanced datasets indicate the effectiveness of the proposed incremental L2MM for imbalanced classification tasks.

Over-treatments such as high medicine fees, repeated physical examinations, overuse of antibiotics and hormones are very common in some hospitals in China. In order to reduce these over-treatment phenomena, two multidimensional data analysis methods are used to analyze historical medical data in this paper. Then, for different situations of outpatients and inpatients, the best treatment cases for recommendation and the worst treatment cases to avoid are found out by using two algorithms based on the Skyline operator to sort multidimensional historical medical data. The results provide patients with appropriate prescriptions and provide doctors with reasonable recommendations of treatments.

Actual dwell time (DT) at rail transit stations is one of the most significant and useful parameters in evaluating the scheduled timetable and making any dynamic operation adjustments as ne...

The unpredicted and unbalanced nature of passenger flow over temporal and spatial complicates the design and optimization of scheduled timetables. This study presents a model based on time-driven microscopic simulation to evaluate train timetable from the viewpoint of big passengers' data on rail transit lines. By calculating the load factor of rail lines for different times and sections, the proposed model assesses the drawbacks of scheduled timetables. It incorporates the number of alighting and boarding passengers, the train load factor, the number of passengers waiting for trains due to overcrowding in vehicles and the number of waiting passengers on the platform into the design. This methodology is examined in a real rail transit line that involves 634 trains and more than 1 million passenger trips. Experimental results demonstrate the efficiency of the proposed model. A time-driven microscopic simulation model to evaluate train timetable with big passengers' data is presented.The drawbacks of scheduled timetables by the analysis of the interaction between trains and passengers is assessed.The train load factors and the extra passengers' waiting times due to overcrowded vehicles are taken into account.A real rail transit line that involves 634 trains and more than 1 million passenger trips per day is examined.

In this paper, we explore the impact of customer impatience on the performance of a production service system that consists of one production inventory subsystem and one service subsystem. It is inevitable that the un-storability of service makes customers wait in the production service system. Waiting for service easily leads to impatient behaviors. The impact of two kinds of impatient behaviors balk and reneging on the production service system performance is examined. The problem is formulated as a Markov system, then the stationary probability of the system is derived to conduct the performance evaluation. The results show that customer impatience has the different impact patterns on the two subsystems (i.e. production inventory subsystem and service subsystem) of the production service system. Specifically, customer impatience leads to the performance improvement of one subsystem along with the performance deterioration of the other subsystem. Meanwhile, service capacity plays an important role in th...

Focus on scheduling of an LP-OKP model with machines availability constraints.Propose slack-based mixed integer linear programming scheduling algorithm.Propose slot-based search scheduling algorithm.Evaluate the proposed algorithms through a real case of power plant building industry. This paper addresses a job shop scheduling problem, in which the jobs belong to the large-piece one-of-a-kind production (LP-OKP) model, and the machines are subject to availability constraints (Jm(LP-OKP), hjk |r-a |Cmax). The objective is considered to minimize the makespan. The scheduling model is derived from real industries, such as ship building industry, power plant building industry, steel-structure building industry and so on, and represents a more realistic and complex situation. In this study, two different scheduling algorithms of a slack-based mixed integer linear programming (MILP) model and a time slot-based search method are proposed to solve the scheduling problem. The experiments are implemented on CPLEX and search scheduling platform, respectively. The results reveal that the slack-based MILP model and time slot-based search method give better solutions than the rule-based algorithm. The deviation of objective optimal value results show that the slack-based MILP model and time slot-based search method can improve the makespan solution by 0.7% and 5.6%, respectively. And the percentage improvement of CPU time results show that they can shorten the CPU time by 74.5% and 12.9%, respectively. We found that the performance of the time slot-based search method is significantly competitive as compared to the slack-based MILP model but its CPU time is significantly higher than that of the slack-based MILP model.

At present, manufacturers tend to bundle a product with its related services as a product service system, to meet customer needs and achieve competitive advantage. Configuring a product service system involves selecting and combining appropriate product and service components, to satisfy individual customer needs. One crucial step to rapidly configure a product service system is to develop formalised configuration rules, which show the relationships between product service components and customer needs, including those expressed as perception needs. In this study, a rough set-based approach is proposed to acquire configuration rules. First, an information table is built by combining the results of factor analysis and questionnaire survey. Then, customer segmentation is accomplished by clustering. Based on the information table and customer segmentation, a dominance-based rough set approach is used to extract the configuration rules. Finally, a weighbridge is selected for the case study to validate the pro...