Search

Container terminal automation offers many potential benefits, such as increased productivity, reduced cost, and improved safety. Autonomous trucks can lead to more efficient container transport. A novel lane detection method is proposed using score-based generative modeling through stochastic differential equations for image-to-image translation. Image processing techniques are combined with Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Genetic Algorithm (GA) to ensure fast and accurate lane positioning. A robust lane detection method can deal with complicated detection problems in realistic road scenarios. The proposed method is validated by a dataset collected from the port terminals under different environmental conditions; in addition, the robustness of the lane detection method with stochastic noise is tested.

The reliability of logistics processes is most often considered in the context of the correct flow of materials. The information that accompanies the logistic flow is usually not considered a potential source of errors or disruptions to the correct operation of the logistic chain. The article presents literature studies in this field and presents a model for evaluating the logistics process, taking into account the quality of information. An example of process evaluation based on a container terminal was presented.

The purpose of this research is to define the bottlenecks in a port's rail container transport process and simulate the proposed scenarios in a real port using the lean-railroading approach. After visiting the selected port and holding several meetings with the authorities, all functions relating to containers have been identified. Each section's arrival and departure times were used in this analysis for all containers shipped by rail in 2017, including 19,000 records. The data was accomplished by the processing time of transferring the containers between the railway areas and the port berth. The value stream map (VSM) related to the processes was prepared using the lean approach, the current port situation and proposed scenarios were simulated using the AnyLogic software. The results showed that the two proposed solutions effectively reduced time and costs by up to 20%. The research port also could increase the rail share to 8% without spending on infrastructure. Nevertheless, the rail container terminal warehouse will be a significant bottleneck for the port by growing the load operations. Reducing the turnover of the wagon will also halve the required number of wagons in the network. Therefore, it is recommended that a framework be established to direct freight to rail transport through the Ports Organization by meeting the infrastructure and operational requirements of the port and through the Railroad Company by launching the scheduled freight train and providing the required fleets. [ABSTRACT FROM AUTHOR]

Southern Vietnam is a key economic region for the country, producing 45% of Vietnam's GDP, despite only accounting for 9% of the country's total landmass and 21% of its population. Specifically, the Southern Region dominates seaport-based container transportation, contributing 55% of Vietnam's total vessel calls and 74% of its total container cargo volume, with the North and Central regions trailing far behind. However, despite the sector's importance, the efficiency of container terminals in Southern Vietnam has not been evaluated by the literature in recent years. This study fills this research gap by using data envelopment analysis (DEA) to evaluate the efficiency of major container terminals in Southern Vietnam. The results of the DEA Malmquist analysis show that most container terminals demonstrated productivity growth between 2017 and 2019, and that this mainly derived from advancements in catch-up operational efficiency and in technological innovation. Moreover, the DEA slack-based measure and DEA undesirable output analyses demonstrate that three container terminals (Cai Mep International Terminal, Tan Cang Cat Lai, and Tan Cang–Cai Mep International Terminal) achieved the maximum efficiency level when both applying and not applying the undesirable output variable. However, other container terminals displayed severe operational inefficiency, with these terminals' efficiency values decreasing by between 9% and 40% when considering undesirable outputs. The ineffective terminals can therefore be categorized as consuming excessive inputs and lacking in the production of desirable outputs or creating large amounts of undesirable outputs. Therefore, using these results, port authorities and policymakers should be better placed to develop appropriate strategies for managing container terminals in Southern Vietnam. [ABSTRACT FROM AUTHOR]

With increasing volumes of goods, the world economy has never been so dependent on the sea than before. The main concern of developed countries is therefore to guarantee the safety of maritime transport and above all to optimize logistics operations within port sites. Internationally, many countries such as China have been able to distinguish themselves with increasingly efficient ports. Innovating and improving the logistics organization of seaports has become a real challenge for them. Nevertheless, many authors question the management of ports in France. Some speak of the weakness of French ports in terms of containerization. Thus, France must review its port infrastructures in order to compete with other world powers. [ABSTRACT FROM AUTHOR]

When road trucks retrieve inbound containers from a container terminal, retrieval operations are often delayed significantly because of the frequent rehandling. We attempt to reduce the number of rehandling operations of inbound containers by using various information sources related to truck arrivals. The sources include dwell time distribution of inbound containers, truck dispatching notice, truck appointment, and real-time position of trucks. Heuristic algorithms are developed to locate and premarshal inbound containers. A simulation study is carried out to test the heuristic algorithms and evaluate the contributions of each information source to terminal performance measures. Practical data on more than 200,000 inbound containers over a period of 8 months and the specifications of the storage yard and rail mounted gantry cranes (RMGCs) of a container terminal in Busan, South Korea, are used for the simulation study. Our results show that the truck system time and number of rehandling operations during a retrieval are, on average, reduced by 47 and 98%, respectively, by utilizing the above information. The truck-dispatching notice, in particular, has a significant contribution to the reductions. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The paper presents an Integrated Maintenance Decision Making Model (IMDMM) concept for cranes under operation especially into the container type terminals. The target is to improve cranes operational efficiency through minimizing the risk of the Gantry Cranes Inefficiency (GCI) results based on the implementation of the Digital Twins concept for maintenance purposes. The proposed model makes a joint transportation process and crane maintenance scheduling, relevant to assure more robust performances in stochastic environments, as well as to assess and optimize performances at different levels, from components and transport device to production systems (container terminal). The crane operation risk is estimated with a sequential Markov chain Monte Carlo simulation model and the optimization model behind of IMDMM is supported through the Particle Swarm Optimization algorithms because the objective function a non-linear stochastics problem with bounded constrains. The developed model allows the container terminal operators (management process) to obtain a maintenance schedule that minimizes the GCI (holistic indicator), as well as establishing the desired level of risk. The paper demonstrates the effectiveness of the proposed maintenance decision making concept model for cranes under operation using data from of a real container terminal (case study). [ABSTRACT FROM AUTHOR]

: The paper studies two different types of container strategy: the traditional one, which involves container allocation into positions with a minimum stack height; and the “temperature” one, which considers the positions with a minimum “temperature” for the containers in the stack below. The “temperature” in this case means a metric which is connected to the storage (dwell) time of the containers. Particularly, this value can be represented by container selection probability or number of days stored in the container yard. Utilization of different metrics results in significantly different numbers of moves. Therefore, the authors compare two container stacking strategies: the traditional one and the temperature strategy with different metrics. It is stated that these strategies can be compared only by simulation modelling. The paper describes the main algorithms used to provide simulation modelling. The results of the research show that the temperature strategy with container dwell time as temperature metrics can save 6% of the total moves necessary to maintain the container flow. [ABSTRACT FROM AUTHOR]

The paper studies two different types of container strategy: the traditional one, which involves container allocation into positions with a minimum stack height; and the “temperature” one, which considers the positions with a minimum “temperature” for the containers in the stack below. The “temperature” in this case means a metric which is connected to the storage (dwell) time of the containers. Particularly, this value can be represented by container selection probability or number of days stored in the container yard. Utilization of different metrics results in significantly different numbers of moves. Therefore, the authors compare two container stacking strategies: the traditional one and the temperature strategy with different metrics. It is stated that these strategies can be compared only by simulation modelling. The paper describes the main algorithms used to provide simulation modelling. The results of the research show that the temperature strategy with container dwell time as temperature metrics can save 6% of the total moves necessary to maintain the container flow.

Due to mega-ships, increasing container throughput, and nonuniform truck arrivals, many container terminals face challenges of unbalanced workloads of yard equipment, shortage of equipment resources in peak hours, and congestion problem. To solve such issues, we propose a mixed-integer bilevel programming model to optimize the vessel-dependent time windows for inbound trucks and yard crane deployment simultaneously. In the proposed bilevel model, the upper level aims to minimize the total truck waiting time at the container terminal gate and yard, while the lower level is formulated to minimize the total workload overflow to next shift in the whole container yard. The optimal yard crane deployment obtained in the lower level will transfer to the upper level problem to determine the waiting time of trucks in the yard and then affect the truck arrivals pattern. To solve the model, a hybrid algorithm—called hybrid genetic algorithm, based on collective decision optimization—is put forward by combining the genetic algorithm and the collective decision optimization algorithm. Numerical experiments are conducted to validate the proposed approach is effective to simultaneously flatten truck arrivals and improve the efficiency of yard cranes. The proposed approach can significantly reduce container terminals’ truck waiting time.

Introduction: In the paper, we examine the energy consumption efficiency of specialized container diesel trucks engaged in container transportation at a seaport terminal. Objectives: Using the container terminal at Klaipėda in Lithuania as the background for the research, we produced an improved energy consumption model for measuring the theoretical energy consumption and regeneration of diesel trucks at the terminal and provide a comparative analysis. Methods: We created a mathematical model which describes the instantaneous energy consumption of the diesel trucks, taking into account their dynamic properties and the overall geometry of their routes—"Ship-Truck-Stack-Ship"—using the superposition principle. We investigated other critical parameters relevant to the model and provide a statistical evaluation of the transportation process using data from a case study of Klaipėda port, where we collected measurements of container transportation parameters using georeferenced movement detection and logs from wireless equipment positioned on the diesel-powered container trucks. Results: The modeling results showed that an instantaneous evaluation of energy consumption can reveal areas in the container transportation process which have the highest energy loss and require the introduction of new management and process control initiatives to address the regulations which are designed to decrease harmful industrial emissions and encourage novel technologies and thereby increase the eco-friendliness of existing systems. Conclusion: Based on the research results, the article can provide a reference for the estimation of diesel truck efficiency in seaport terminal operations. [ABSTRACT FROM AUTHOR]

New container terminals are embracing robotized transport vehicles such as lift-automated guided vehicles (LAGVs) and automated guided vehicles (AGVs) to enhance the terminal throughput capacity. Although LAGVs have a high container handling time, they require less coordination with other terminal equipment in comparison with AGVs. In contrast, AGVs are hard-coupled resources, require less container handling times, but operate with high coordination delays in comparison with LAGVs. The effect of such operational trade-offs on terminal performance under various design parameter settings, such as yard block layout and a number of resources, is not well understood and needs to be evaluated at the terminal design phase. To analyze these trade-offs, we develop stylized semi-open queuing network models, which consist of two-phase servers and finite capacity queues. We develop a novel network decomposition method for solving the proposed queuing models. The accuracy of the solution method is validated using detailed simulation models. Using the analytical models, we study the performance trade-offs between the transport vehicle choices: LAGVs and AGVs. Our results show that the throughput capacity of the terminal in the container unloading process increases by up to 16% if LAGVs are chosen as transport vehicles instead of AGVs. However, at certain parameter settings, specifically, when the arrival rate of containers is low, the throughput time performance of the terminal is higher (up to 8%) with AGVs than with LAGVs. We also derive insights on the yard block layout and the technology choice for quay cranes. [ABSTRACT FROM AUTHOR]

Berth allocation is one of the crucial points for efficient management of ports. This problem is complex due to all possible combinations for assigning ships to available compatible berths. This paper focuses on solving the Berth Allocation Problem (BAP) by optimising port operations using an innovative model. The problem analysed in this work deals with the Discrete and Dynamic Berth Allocation Problem (DDBAP). We propose a novel mathematical formulation expressed as a Mixed Integer Linear Programming (MILP) for solving the DDBAP. Furthermore, we adapted a metaheuristic solution approach based on the Bee Colony Optimisation (BCO) for solving large-sized combinatorial BAPs. In order to assess the solution performance and efficiency of the proposed model, we introduce a new set of instances based on real data of the Livorno port (Italy), and a comparison between the BCO algorithm and CPLEX in solving the DDBAP is performed. Additionally, the application of the proposed model to a real berth scheduling (Livorno port data) and a comparison with the Ant Colony Optimisation (ACO) metaheuristic are carried out. Results highlight the feasibility of the proposed model and the effectiveness of BCO when compared to both CPLEX and ACO, achieving computation times that ensure a real-time application of the method. [ABSTRACT FROM AUTHOR]

The port sector has played an important role in global trade, with ports acting as a transportation chain-ring in environmental-social performance improvement. The usage of sea transportation means has spread across the world. Starting with the Kyoto Protocol for ships, the environmentally friendly trend has encompassed the port sector. However, it is difficult to find a model with the same characteristics as those of the ports as the models. The models can be used to compare operational performance regarding carbon dioxide (CO2) emission production. This research aimed to estimate CO2 emissions at container ports to portray how a port deals with its operational matters, using models suitable for ideal circumstances based on available equipment. This calculative system applies a bottom-up calculation of the work activities at a port, treating the amount of fuel consumption not as an input variable, but as the result of the calculation itself. The input variables include throughput, transshipment process, transportation modality, and terminal layout. The results show that several equipment operational activities can be optimized by comparing the calculation results for actual CO2 emissions. It was found that each twenty-foot equivalent unit produced as much as 11.27 kg of CO2 emissions at the Belawan International Container Terminal in Medan, Indonesia. This research has considerable potential use for ports, showing how to calculate CO2 emissions at a port under ideal circumstances, that models in use can be adapted to any port characteristics, and that the data serving as the input variables are not difficult to obtain.

This paper analyses changes of berth infrastructure and suprastructure by global container terminals (CTs) and by four eastern Adriatic ports in the last decade. The emphasis is on understanding whether CTs at Koper, Trieste, Rijeka and Bar achieved higher berth utilisation and productivity per ship-to-shore (STS) crane and if so, how and whether their development is in line with the global trend in CT berth productivity. On this basis a comparison model of twenty selected global CTs is used for productivity comparison as a first step in the process of analysing subsystem productivity. The study shows that four eastern Adriatic ports made different decisions, but with the same goals in reaction to the increased flow of containers via the Adriatic Sea transport route. Their main goal was to increase berth productivity by controlling the eventual subsystem overcapacity. According to observations, the Port of Koper is running at the subsystem’s upper level, while CTs in Trieste, Rijeka and Bar operate with certain degree of berth infrastructural, and suprastructural overcapacity.

During the past two decades, there has been an increase on maritime freight traffic particularly in container flow. Thus, the Berth Allocation Problem (BAP) can be considered among the primary optimization problems encountered in port terminals. In this paper, we address the Dynamic Berth Allocation Problem (DBAP) in a conventional layout terminal which differs from the popular discrete layout terminal in that each berth can serve multiple vessels simultaneously if their total length is equal or less than the berth length. Then, a Modified Sailfish Optimizer (MSFO) meta-heuristic based on hunting sailfish behavior is developed as an alternative for solving this problem. Finally, computational experiments and comparisons are presented to show the efficiency of our method against other methods presented in the literature in one hand. We also discuss the productivity of a container terminal based on different scenarios which can happen.

Storage space allocation and handling operation problems are two main crucial problems in container terminals. Early research efforts, however, are seldom devoted to studying them together. Therefore, this paper considers these two problems simultaneously for outbound containers in rail-water intermodal container terminals (RWICTs), where rail-mounted gantry cranes, inner trucks, and quay cranes are involved. A two-stage problem is proposed: Stage 1 is to determine locations of the containers and reduce the overlapping amount, considering container weight, departure time, destination ports, and containers left from earlier planning periods in railway container yards, according to the locations of containers from Stage 1; and Stage 2 aims at obtaining optimal job sequences of different types of equipment and minimizing makespans of handling operations, considering some operational constraints, particularly rehandling time and inner truck congestion. To solve the problem, a two-stage heuristic algorithm is proposed, where the rolling planning horizon and a new update strategy are introduced. A heuristic algorithm is introduced in Stage 1 and a novel two-layer genetic algorithm is proposed in Stage 2, which introduces proximity principles and the reselection operation. Afterward, the results from Stage 2 are used to resolve the first stage problem, while the problem in Stage 2 is also resolved using the new results from Stage 1. This iterative process continues until there are no more improvements in Stage 1. Finally, the results of the computational experiments indicate that the proposed model and solution approaches are effective and efficient in solving the two-stage problem for outbound containers in RWICTs.

In vehicle routing problems, time windows are often used to formulate partial order relations between tasks in optimization model and related algorithm. However, time-window and partial order constraints are not equivalent. In the context of stacking operations in container terminals, this study proposed a relaxation method to translate the time-window constraints into the partial order constraints and built a routing optimization model with the partial order constraints. Under the framework of branch and bound (BB) algorithm, this paper studied the relationship between time-window constraints and partial order constraints. The features of the two models are analyzed. The Solomon datasets are used to conduct numerical analysis and examine the differences between the two models in terms of optimization ability and optimality. Revealed by experimental results, the partial-order-based model can be solved with better optimality, while the time-window-based mode can be solved within a shorter time. Moreover, inspired by the effects of narrowing the time windows, it will be a new research direction to develop improved algorithms for partial-order-based model based on time-window-based decomposition method.

This work is devoted to the study of the Uncertain Quay Crane Scheduling Problem (QCSP), where the loading /unloading times of containers and travel time of quay cranes are considered uncertain. The problem is solved with a Simulation Optimization approach which takes advantage of the great possibilities offered by the simulation to model the real details of the problem and the capacity of the optimization to find solutions with good quality. An Ant Colony Optimization (ACO) meta-heuristic hybridized with a Variable Neighborhood Descent (VND) local search is proposed to determine the assignments of tasks to quay cranes and the sequences of executions of tasks on each crane. Simulation is used inside the optimization algorithm to generate scenarios in agreement with the probabilities of the distributions of the uncertain parameters, thus, we carry out stochastic evaluations of the solutions found by each ant. The proposed optimization algorithm is tested first for the deterministic case on several well-known benchmark instances. Then, in the stochastic case, since no other work studied exactly the same problem with the same assumptions, the Simulation Optimization approach is compared with the deterministic version. The experimental results show that the optimization algorithm is competitive as compared to the existing methods and that the solutions found by the Simulation Optimization approach are more robust than those found by the optimization algorithm.

In the existing literature, the berthing operations, the quay crane assignments, and the scheduling problems were usually conducted without considering the worker performances (WPs) and the yard truck tasks (YTTs). However, professionals in situ corroborate the crucial effect of WPs and the yard YTTs on quay crane performance and efficiency. This study introduced a new feasible model for investigating the berth and the quay crane assignments based on the scheduling problem, including worker performances and yard truck deployment constraints. First, a mixed-integer programming (MIP) model is implemented to reduce the vessel’s departure time. Then, a particle swarm optimization (PSO) algorithm is introduced to solve the problems. The Dar es Salaam port is selected as a case study to test the proposed model with a real-time dataset that was collected from a multinational company managing container terminals. The results show the efficiency and the accuracy of the proposed model. The PSO algorithm is 86% and 62% more time-saving than MILP and T2S solutions for a small number of containers, respectively. Additionally, the PSO solution is 73% and 53% time-saving for a medium number of containers than MILP and T2S models, respectively. Finally, the present study proposes consideration of the worker assignment and the yard truck deployment during the planning phase.

Background: Cargo-handling equipment (CHE) plays a vital role in maintaining the efficiency of a highly-worked container terminal. Methods: This study is aimed to analyze the CHEs’ performance, which is conducted based on a contextual application of the overall equipment-effectiveness (OEE) technique and data collected from a field survey in 14 container terminals in Vietnam. Results: The findings reveal that the CHEs are operated incompatibly with their actual capacity due to low performance. Also, the findings clarified the unproductive exploitation of container terminals and low actual terminal throughput since the capacity-designed terminals are currently operating above their actual capacity. Conclusions: The application of the OEE index for groups of CHE equipment is an origin for the impact assessment of the overall performance between the groups of CHEs’ equipment, thereby proposing management tools for supporting the improvement of the CHEs and container terminals’ performance in Vietnam.

Background: Scheduling the arrival of external trucks in container terminals is a critical operational decision that faces both terminal managers and trucking companies. This issue is crucial for both stakeholders since the random arrival of trucks causes congestion in the terminals and extended delays for the trucks. The objective of scheduling external truck appointments is not only to control the workload inside the terminal and the costs resulting from the excessive waiting times of trucks but also, to reduce the truck turnaround time. Methods: A binary programming model was proposed to minimize the waiting time cost, demurrage cost, and container delivery cost. Moreover, a sensitivity analysis was performed to compare various scenarios in terms of cost and to study to what extent the workload level is affected. The mathematical model was solved using Gurobi© 8.1.0 software. Results: 30 instances found in the literature were solved and evaluated in terms of the objective function value (i.e., cost) and truck turnaround time before and after controlling the workload inside the container terminal using the new proposed constraint. Conclusions: The obtained results showed a better distribution of the terminal workload, as well as a lower truck turnaround time that reduces the total cost.

This study considers the Dynamic Minimum Cost Hybrid Berth Allocation Problem (DMCHBAP) with fixed handling times of vessels. The objective function to be minimized consists of three components: costs of positioning, waiting, and tardiness of completion for all vessels. A mathematical formulation of DMCHBAP, based on Mixed Integer Linear Programming (MILP), is proposed and used within the framework of commercial CPLEX 12.3 solver. As the speed of finding high-quality solutions is of crucial importance for an efficient and reliable decision support system in container terminal, two population-based metaheuristic approaches to DMCHBAP are proposed: combined Genetic Algorithm (cGA) and improvement-based Bee Colony Optimization (BCOi). Both cGA and BCOi are evaluated and compared against each other and against state-of-the-art solution methods for DMCHBAP on five sets of problem instances. The conducted computational experiments and statistical analysis indicate that population-based metaheuristic methods represent promising approaches for DMCHBAP and similar problems in maritime transportation. [ABSTRACT FROM AUTHOR]

In recent years, the bottleneck of yard operation in container terminals has become more and more prominent. For resolving the departure delay problem of vessels, this paper proposes a mixed integer programming model considering the operation time requirements of vessels in a container terminal. It can optimize the yard template plan by reasonably arranging the number of loading clusters of each vessel in the yard, and meanwhile, enhance the efficiency of management in yard and lower the departure delay of vessels. Moreover, a case study is presented to illustrate the proposed model and sensitivity analysis is conducted in the numerical experiments to verify the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

Reshuffling containers, one of the daily operations at a container terminal, is time consuming and increases a ship’s berthing time. We propose a decision-tree heuristic to minimise the expected number of reshuffles when arriving containers should be stacked in a block of containers with an arbitrary number of piles. The heuristic algorithm uses the optimal solutions of a stochastic dynamic programming model. Since the total number of states of the dynamic programming model increases exponentially, the model can only solve small-scale problems in a reasonable time. To solve large-scale problems, the heuristic uses the results of the exact model for small-scale problems to generate generalised decision trees. These trees can be used to solve problems with a realistic number of piles. The numerical experiments show the effectiveness of the algorithm. For small-scale problems, the trees can quickly make optimal decisions. For large-scale problems, the decision-tree heuristic significantly outperforms stacking policies commonly used in practice. Using the decision trees, we can compare the performance of a shared-stacking policy, which allows containers of multiple ships to be stacked on top of each other, with a dedicated-stacking policy. Shared-stacking appears to outperform dedicated-stacking. [ABSTRACT FROM PUBLISHER]

Congestion at terminal gates is a significant issue that many ports are dealing with. The port of Songkhla, which serves as the main seaport in southern Thailand, has been confronted with this issue, together with an increase in the number of trucks over the last five years. This study aims to investigate alternatives for improving service operations at the terminal gate, including gate lane expansion, queueing policy in exit-gate sharing, and work process reorganization. A simulation model was developed to analyze the queueing performance: truck queue length and gate waiting time. Scenarios analysis was performed and compared to the baseline model to investigate the impact of proposed approaches on gate performance. The simulation results confirm that a gate-sharing policy can alleviate congestion at a certain truck arrival rate. The method of reorganizing work processes by separating customs work from gate operations and placing it in a new work station appears most appealing. It produces a significant improvement in congestion, comparable to the performance of the gate lane expansion approach while requiring less investment.

As more and more container terminals are becoming intelligent, different kinds of sensors are widely installed at different locations of the cranes and collect a large amount of data. In order to effectively utilize and manage these huge amounts of actual working data of different sensors and grasp the status of the terminal, this article proposes a data processing framework that integrates the crane load, energy consumption, crane trolley speed and crane gearbox vibration signals of the container terminal. Firstly, the load spectrum of the crane load is calculated by the non-parametric density estimation method in probabilistic statistics and the energy consumption curves are obtained. Secondly, the driving cycle of the crane trolley speed are constructed by drawing on the method in the transportation field. Finally, the vibration signals of the crane gearbox are used for anomaly detection by unsupervised methods; at the same time, clustering results can also be used as the basis for extracting typical vibration signals and removing redundant data.