Dynamic Optimization Method Research Articles

Computation Resource Optimization for Large-Scale Intelligent Urban Rail Transit: A Mean-Field Game Approach

Offloading tasks in smart devices (SDs) to an edge intelligence-empowered service centre (EISC) is a promising solution to support burgeoning intelligent applications in large-scale intelligent urban rail transits (URTs). However, dynamic computation resource allocation is still a crucial challenge, facing that various large-scale SDs share the EISC computation resources and the reality that the allocated computation resource for an SD is coupled with the offloading rate of all SDs. This paper proposes a joint dynamic offloading rate control and computation resource optimization method for large-scale intelligent URTs. Firstly, we model the large-scale multi-agent computation resource competition problem by a multi-player differential game (MPDG) and prove that the Nash equilibrium (NE) based optimal solution exists for each SDs. Then, we transform the MPDG model into a mean-field game (MFG). By introducing the mean-field into the game, we can solve the multi-agent optimization problem with a single-agent optimization method. We illustrate the rationality of the MFG model and propose an iterative solution method based on the finite difference method to derive the solution. Finally, we propose a z-transforming-based control method to dynamically reschedule computation resources among intelligent applications to achieve a satisfactory quality of service (QoS). Extensive simulation results show that our proposed scheme can significantly improve the performance of intelligent URTs.

Multi-Agent Reinforcement Learning Method for Disassembly Sequential Task Optimization Based on Human–Robot Collaborative Disassembly in Electric Vehicle Battery Recycling

Abstract With the wide application of new Electric Vehicle (EV) batteries in various industrial fields, it is important to establish a systematic intelligent battery recycling system that can be used to find out the resource wastes and environmental impacts of the retired EV battery. By combining the uncertain and dynamic disassembly and echelon utilization of EV battery recycling in the remanufacturing fields, human–robot collaboration (HRC) disassembly method can be used to solve huge challenges about the efficiency of retired EV battery recycling. In order to find out the disassembly task planning based on HRC disassembly process for retired EV battery recycling, a dynamic disassembly sequential task optimization method algorithm is proposed by Multi-Agent Reinforcement Learning (MARL). Furthermore, it is necessary to disassemble the retired EV battery disassembly trajectory based on the HRC disassembly task in 2D planar, which can be used to acquire the optimal disassembly paths in the same disassembly planar by combining the Q-learning algorithm. The disassembly task sequence can be completed through standard trajectory matching. Finally, the feasibility of the proposed method is verified by disassembly operations for a specific battery module case.

Service-oriented multi-skilled technician routing and scheduling problem for medical equipment maintenance with sudden breakdown

Motivated by the service-oriented intelligent maintenance activity under the development of the digital economy, we study the operation and maintenance strategies of the Maintenance Scheduling and Technician Routing Problem (MSTRP) for medical equipment implemented by multi-skilled technicians. We consider the maintenance services provided for both preventive maintenance (PM, e.g., periodic maintenance based on the contract schedule) and corrective maintenance (CM, e.g., repair action in the event of a sudden breakdown). As these activities are performed on equipment with various types of requirements at different geographically distributed systems, technicians with diverse qualifications need to build teams and visit the matching maintenance requests. Meanwhile, we consider the maintenance actions concerning the dynamic change in equipment failure rate and establish a flexible maintenance mechanism by balancing the total cost. This paper develops an efficient hybrid heuristic based on a genetic algorithm, variable neighborhood search, and other optimization strategies to solve this multi-period equipment maintenance scheduling and multi-skilled technician routing and scheduling problem. Based on the settings of the studied MSTRP, this paper designs different sizes of testing instances and multi-period explanatory cases before conducting extensive comparative experiments and sensitivity analysis to demonstrate the effectiveness of our proposed dynamic maintenance strategy and optimization method.

Multi-Time Interval Dynamic Optimization Model of New Energy Output Based on Multi-Energy Storage Coordination

In response to the problem of mismatch between new energy output and multi-energy load requirement in multi-energy power systems, this article proposes a dynamic optimization model for new energy output in multiple time intervals based on multi-energy storage coordination. First, considering the energy conversion characteristics of multi-energy storage, the dynamic optimization method of new energy output based on the discrete division of subinterval of scheduling time is studied. Then, considering the cost of adjusting various resources comprehensively, the optimization objective of new energy output is studied, and a model-solving method based on a directed graph topology distributed algorithm is proposed. Finally, simulation verification was conducted, and the simulation results showed that the method proposed in this paper can effectively suppress the new energy fluctuation and reduce peak-shaving costs.

Triceps surae muscle forces during dynamic exercises in patients with Achilles tendinopathy: A cross-sectional study.

The aim of this study was to investigate the individual triceps surae muscle forces during the execution of six different functional movements and rehabilitation exercises in patients with Achilles tendinopathy compared to a control group. Triceps surae muscle forces of 15 participants with Achilles tendinopathy (AT) and 15 healthy controls were estimated through a combination of experimental data and musculo-skeletal modeling. Three-dimensional motion capture and force plates were used to collect the ankle and knee joint angles and moments during three functional movements (walking, heel walking, and toe walking) and three rehabilitation exercises (bilateral heel drop, unilateral heel drop with extended knee and with flexed knee). A dynamic optimization method was used to obtain the modeled triceps surae muscle forces. Force-sharing strategies were calculated at the peak triceps surae muscle force and compared between groups. Lower peak triceps surae forces were obtained for the AT group during dynamic exercises. Across all exercises, the average contribution of the soleus (SOL) to the total triceps surae muscle force was the largest (60.83 ± 13.89% [AT] > 56.90 ± 16.18% [healthy]), followed by the gastrocnemius medialis (29.87 ± 10.67% [AT] < 32.19 ± 12.90% [healthy]) and the gastrocnemius lateralis (9.30 ± 4.31% [AT] < 10.91 ± 4.66% [healthy]). The triceps surae force-sharing strategy was different for the toe walking, heel walking, and the bilateral and unilateral heel drop with extended knee. This study provides evidence for altered triceps surae muscle force-sharing strategies during dynamic tasks in patients with AT. The influence of altered muscle force-sharing on the subtendon nonuniformity and/or the tendon loading should be explored in future work.

Research into Dynamic Error Optimization Method of Impeller Blade Machining Based on Digital–Twin Technology

A TC4 impeller blade is a typical weak, rigid, thin–walled part. The contact area between a cutting tool and a workpiece has strong time–varying characteristics. This leads to a strong non–linear variation in cutting load. So, in this kind of part, the processing error is difficult to control. To solve this problem, a method of processing error prediction and intelligent controlling which considers the effect of tool wear time variation is proposed by combining digital–twinning technology. Firstly, an iterative model for digital–twin process optimization is constructed. Secondly, an iterative prediction model of the machining position following the milling force and considering the effect of tool wear is proposed. Based on these models, the machining error of the TC4 impeller blade under dynamic load is predicted. Dynamic machining error prediction and intelligent control are realized by combining the digital–twin model and the multi–objective process algorithm. Finally, the machining error optimization effect of the proposed digital–twin model is verified via a comparison experiment of impeller blade milling. In terms of the precision of milling force mapping, the average error after optimization is less than 8%. The maximum error is no more than 14%. In terms of the optimization effect, the average error of the optimized workpiece contour is reduced by about 20%. The peak contour error is reduced by approximately 35%.

Device–Edge Collaborative Differentiated Data Caching Strategy Toward AIoT

Caching AIoT data at the edge can reduce the load on cloud centers while providing real-time services for AIoT users. Existing static caching strategies based on popularity prediction fail to meet users’ demands for time-varying entity data, while dynamic caching strategies focus only on evaluating the time-varying state characteristics of entity data, but ignore the differences in popularity among entities, resulting in poor service experience. To this end, a device-edge collaborative differentiated data caching strategy considering static entity popularity as well as dynamic system state is proposed. First, a hot entity recognition method centered on user interests is designed to achieve individual preference estimation by mining users’ long short-term interests, and then achieve group interest prediction based on social computing. Based on this, a dynamic caching optimization method is designed, which considers the timeliness of entity data and communication cost of system to design the objective function of optimal cache decision and then solve it based on reinforcement learning. Simulation results demonstrate that the proposed caching strategy achieves better performance than other benchmark strategies in terms of cache hit rate and search cost.

Analyzing the Impact of Depreciation-Estimating Methods on State Transportation Agencies’ Equipment Replacement Decisions Using Dynamic Programming

Replacing equipment at the most economical time not only helps to save state transportation agencies (STAs) costs for operating the fleet but also keeps the fleet’s level of service at an optimal level. Prior research studies focused on developing alternative economic-oriented equipment replacement models rather than the equivalent annual cost (EAC) model to achieve better economic decisions. In addition, various optimization techniques were applied to equipment replacement problems with different objectives, constraints, and contexts. However, few studies examined the impact of depreciation estimation on the equipment replacement decision within STAs by minimizing total equipment cost over a finite study period using the dynamic programming optimization method. This study performed a case study of two class codes of equipment [1.5 m3 (2-yd) diesel engine front-end loaders and 0.453 t (half-ton) fleetside pickup trucks] to analyze the impact of different depreciation calculations on equipment replacement decisions. Using real-world data provided by the Oklahoma Department of Transportation, the study showed that the double-declining balance depreciation method substantially reduces the number of pieces of equipment recommended for replacement compared with the result of the straight-line depreciation method. This study contributes to the understanding of the impact of depreciation methods on equipment replacement decisions as well as the importance of properly estimating equipment depreciation to minimize the equipment costs over a designated study period among STA communities. The demonstrated manual calculation procedures of dynamic programming for cost optimization potentially may facilitate STA’s adoption of dynamic programming for equipment economic decisions.

The real-time dynamic online feeder service with a maximum headway at mandatory stops

On the one hand, fully flexible demand-responsive feeder services efficiently tailor their service to passengers' needs. Traditional services, on the other hand, offer predictability and easier cost control. This paper considers a semi-flexible feeder service that combines positive characteristics of both traditional and fully flexible services. There are two types of bus stops in this service. Mandatory bus stops have a maximum allowable headway for bus departures. Optional stops are only visited when there is demand for transportation nearby. When new passenger requests arrive, the performance of this feeder service is optimised in real time. A metaheuristic with two phases is developed to optimise the service. The dynamic optimisation method is compared to a model that optimises the service when all requests are known beforehand. The results show that the dynamic method has an average gap of 6.5% with respect to the static model and an average acceptance rate of 95.1%. A case study in the city of Antwerp shows that, when compared to existing transit options in the region, this feeder service can increase the service quality by 31.6% when enough resources are available.

Dynamic topology optimization of continuum structures considering moving mass excitations

The structures excited by the moving mass are widely existed in nature and engineering. The dynamic loads are related to time and space and greatly affect structural performances. However, the moving mass problems have not been considered in the topology optimization design of continuum structures. Therefore, this paper proposes a novel dynamic topology optimization method to design the continuum structures considering moving mass excitations. Firstly, the finite element formulation is established to solve the dynamic responses for the moving mass excitations by using the HHT-α method. Secondly, in the density-based framework, the dynamic topology optimization problem is formulated to minimize the mean strain energy constrained with the total volume, whose interpolation scheme is the rational approximation of material properties. Thirdly, the sensitivities of the optimization problem are derived by using the adjoint variable method in the discretize-then-differentiate scheme. Finally, numerical examples demonstrate the effectiveness of the proposed method and the influence of the moving mass excitations on the topological structures.

Incorporation of AIS data-based machine learning into unsupervised route planning for maritime autonomous surface ships

Maritime Autonomous Surface Ships (MASS) are deemed as the future of maritime transport. Although showing attractiveness in terms of the solutions to emerging challenges such as carbon emission and insufficient labor caused by black swan events such as COVID-19, the applications of MASS have revealed problems in practice, among which MASS navigation safety presents a prioritized concern. To ensure safety, rational route planning for MASS is evident as the most critical step to avoiding any relevant collision accidents. This paper aims to develop a holistic framework for the unsupervised route planning of MASS using machine learning methods based on Automatic Identification System (AIS) data, including the coherent steps of new feature measurement, pattern extraction, and route planning algorithms. Historical AIS data from manned ships are trained to extract and generate movement patterns. The route planning for MASS is derived from the movement patterns according to a dynamic optimization method and a feature extraction algorithm. Numerical experiments are constructed on real AIS data to demonstrate the effectiveness of the proposed method in solving the route planning for different types of MASS.

Dynamic schedule model and algorithm optimization method for linear engineering projects

Linear engineering projects require careful consideration of multiple construction scenarios. These scenarios include linear, bar and block activities, with the direction and velocity of linear activities varying, as well as the number of construction teams involved. Resource constraints further complicate scheduling. To address these challenges, a two-stage dynamic scheduling model was developed in this study. The first stage involved selecting a scheduling strategy based on available resources to meet the project deadline, while the second involved developing two cost-effective models to determine the best plan. To optimize the model, a hybrid algorithm incorporating an elite strategy, an adaptive operator and a catastrophe operator was established. The capabilities of the proposed model were demonstrated through several application examples, highlighting the benefits of improved consideration of construction scenarios in linear engineering. Overall, the model’s ability to identify a reasonable range of resource allocations highlights its effectiveness in optimizing linear engineering project schedules.

A Dynamic Task Assignment Optimization Method for Multi-AGV System Based on Genetic Algorithm

&lt;p&gt;Aiming at the influence of AGV without considering the working state on task assignment decision in multi-AGV system task assignment, a dynamic task assignment decision method with task completion prediction based on genetic algorithm. When assigning the arrived tasks at each stage, the decision method brings the working AGVs and the idle AGVs into the set of schedulable vehicles at the same time, which expands the scope of the optimal decision, makes the available AGV resources more fully mobilized in the dynamic scheduling process, and improves the efficiency of the whole scheduling system. First, this paper establishes a prediction model for task completion. On this basis, the task assignment decision model of multi-AGV system based on task completion prediction is established, and the coding, fitness function and genetic operation of the genetic algorithm suitable for this problem are designed. Finally, a univariate factor analysis is carried out on the task assignment time interval and the number of AGVs by using an example, which verifies the effectiveness of the task assignment strategy of the multi-AGV system based on task completion prediction. The results show that the genetic algorithm can better solve the task assignment problem with task completion prediction, and can schedule the available AGV resources to a greater extent, which effectively increase the number of tasks completed by the multi-AGV system in one production cycle.&lt;/p&gt; &lt;p&gt;&amp;nbsp;&lt;/p&gt;

A new optimization method of energy consumption for dynamic boil-off gas

Related research on refrigerant optimization based on the dynamic BOG has yet to be proposed. This study offers a new dynamic optimization method for boil-off gas (BOG) re-liquefaction system to improve the system's energy efficiency. The performance of C3MR (the propane precooling mixed refrigerant cycle) is analyzed based on the steady state model of leading liquefaction equipment in the process simulation in Aspen HYSYS software. And the principle of the energy consumption variation law of propane precooled compressor and mixed refrigerant compressor, based on the actual possible disturbance range obtained in the natural gas liquefaction process, is applied to construct the multi-objective optimization model with minimum energy consumption and maximum heat exchanger efficiency. Finally, the dynamic responses of disturbances were achieved and discussed with the aid of GA (Genetic Algorithm) and PSO (particle swarm optimization). The optimized performance of the dynamic liquefaction cycle is 34.8% better than the steady condition system in terms of energy efficiency, and the maximum energy consumption can be reduced by 50.60%. The advantage of the proposed optimization framework is its adaptability to other dynamic liquefaction processes. It has practical reference significance for the cost control of the BOG re-liquefaction recovery project.

A Dynamic Continuous Constrained Phase Factor Graph Optimization Method for GNSS Kinematic Precise Point Positioning

Carrier phase measurements play a more critical role in achieving precise positioning for lower noise and insensitivity to multipath error over code (pseudorange) measurements. Real-time kinematic (RTK) is the most used precise positioning technique. However, the application costs are increased due to the demands for differential base stations. The undifferenced precise point positioning (PPP) does not require differential station assistance but is typically for static use. This paper proposes a dynamic continuous constrained phase (DCCP) method based on factor graph optimization for kinematic positioning without differential stations, even in the presence of cycle slips. The precise velocity estimated via integral doppler and the time correlation of phase ambiguity are regarded as the probability constraints in the factor graph. A real-world experiment was carried out to verify the feasibility of the DCCP method. Experimental results show that the proposed method had a better performance than the kinematic PPP method and window carrier phase (WCP) optimization method with 0.034m standard deviation and 1.338m mean error. It is our hope that this approach provides new insight into the study of other time-correlated state estimation problems.

A dynamic emergency planning system for relocating vulnerable people to safe shelters in response to heat waves

A significant proportion of the world's population is ageing. As a result, numerous authorities and governments are formulating strategies to improve their citizens' services. Nevertheless, climate change and natural hazards have revealed the need for additional emergency plans tailored to the needs of the elderly. While recent studies have focused on certain types of disasters, such as floods and earthquakes, heat waves have not received as adequate attention in disaster response research. Using a model of emergency response developed in this study, we propose a dynamic optimization method to facilitate the rapid relocation of vulnerable people to designated shelters in order to minimize their exposure to heat waves. The proposed model views the problem as a dynamic evacuation planning in which requests for assistance from the emergency centre are received after the collection of people from their homes has begun. Subsequently, using some efficient optimisation methods, the strategy gets updated on a regular basis. The proposed model is evaluated using a case study that takes into account a randomly generated set of locations and relocation request in a local government area in Sydney, Australia. Results indicate that the proposed method is capable of generating acceptable response plans for relocating elderly individuals in the shortest possible time during actual emergencies induced by heat waves. In terms of emergency preparation, administration, and policy, this study's findings have implications for countries with urban areas that are particularly vulnerable to climate change effects.

Ocean data compression based on block SVD

Fast development of ocean observations and numerical modeling increases the need for data transmission, storage and extraction. This paper presented a new data compression method based on Singular Value Decomposition (SVD) with data matrix divided into different sub-matrices with consideration of odevity and remnant. An automatic matrix-dividing method is applied to divide smartly the data matrix into sub-matrices. These sub-matrices are then compressed based on an improved SVD, which enhances the compression performance by utilizing the orthogonal property of vectors generated by SVD. A dynamic optimization method which is capable of determining the proper scale of retained data under the accuracy requirement of ocean data is also established. Two indices are derived mathematically to search the best block pattern quickly. The performance and reliability of the block-based SVD compression is verified with the successful compression and recovery of the Hybrid Coordinate Ocean Model data.

A Nonsmooth Dynamical Systems Perspective on Accelerated Extensions of ADMM

Recently, there has been great interest in connections between continuous-time dynamical systems and optimization methods, notably in the context of accelerated methods for smooth and unconstrained problems. In this paper we extend this perspective to nonsmooth and constrained problems by obtaining differential inclusions associated to novel accelerated variants of the alternating direction method of multipliers (ADMM). Through a Lyapunov analysis, we derive rates of convergence for these dynamical systems in different settings that illustrate an interesting tradeoff between decaying versus constant damping strategies. We also obtain modified equations capturing fine-grained details of these methods, which have improved stability and preserve the leading order convergence rates. An extension to general nonlinear equality and inequality constraints in connection with singular perturbation theory is provided.

Control strategy for dynamic operation of multiple chillers under random load constraints

Under the influence of building load stochasticity and thermal inertia of central air conditioning system, the lag of control strategy makes it difficult for the cooling source system to meet the cooling demand of users in real time, which affects the operational energy efficiency ratio of the cooling source system. In this regard, this paper proposes a chiller dynamic optimization method based on Markov chain model predictive control algorithm, which studies the optimization of parallel operation characteristics of multiple chillers on the basis of correcting the occurrence of small probability events and avoiding frequent chiller starts and stops, reveals the prediction law of hourly load distribution of the future air conditioning system, explores the balance point between the prediction time of building load and the calculation time of chiller control strategy, provides a theoretical basis and reference for determining the time limit of the optimal calculation and the output of the optimal control strategy. The study provides a theoretical basis and reference for determining the time limit of the optimal calculation and the output of the optimal control strategy. It is found that the optimal control strategy not only has good optimization effect, but also the confidence is significantly increased by 56.7%.

Dynamic Path Optimization Based on Improved Ant Colony Algorithm

Dynamic path optimization is an important part of intelligent transportation systems (ITSs). Aiming at the shortcomings of the current dynamic path optimization method, the improved ant colony algorithm was used to optimize the dynamic path. Through the actual investigation and analysis, the influencing factors of the multiobjective planning model were determined. The ant colony algorithm was improved by using the analytic hierarchy process (AHP) to transform path length, travel time, and traffic flow into the comprehensive weight-influencing factor. Meanwhile, directional guidance and dynamic optimization were introduced to the improved ant colony algorithm. In the simulated road network, the length of the optimal path obtained by the improved ant colony algorithm in the simulation road network is 3.015, which is longer than the length of the optimal path obtained by the basic ant colony algorithm (2.902). The travel time of the optimal path obtained by the improved ant colony algorithm (376 s) is significantly shorter than that of the basic ant colony algorithm (416.3 s). The number of iterations of the improved ant colony algorithm (45) is less than that of the basic ant colony algorithm (58). In the instance network, the number of iterations of the improved ant colony algorithm (18) is less than that of the basic ant colony algorithm (26). The travel time of the optimal path obtained by the improved ant colony algorithm (377.1 s) is significantly shorter than that of the basic ant colony algorithm (426 s) and the spatial shortest distance algorithm (424 s). Compared with the basic ant colony algorithm and the spatial shortest distance algorithm, the results of the optimal path obtained by the improved ant colony algorithm were more accurate, and the effectiveness of the improved ant colony algorithm was verified.