Multiple Local Search Research Articles

Robotic navigation with deep reinforcement learning in transthoracic echocardiography

Abstract Purpose The search for heart components in robotic transthoracic echocardiography is a time-consuming process. This paper proposes an optimized robotic navigation system for heart components using deep reinforcement learning to achieve an efficient and effective search technique for heart components. Method The proposed method introduces (i) an optimized search behavior generation algorithm that avoids multiple local solutions and searches for the optimal solution and (ii) an optimized path generation algorithm that minimizes the search path, thereby realizing short search times. Results The mitral valve search with the proposed method reaches the optimal solution with a probability of 74.4%, the mitral valve confidence loss rate when the local solution stops is 16.3% on average, and the inspection time with the generated path is 48.6 s on average, which is 56.6% of the time cost of the conventional method. Conclusion The results indicate that the proposed method improves the search efficiency, and the optimal location can be searched in many cases with the proposed method, and the loss rate of the confidence in the mitral valve was low even when a local solution rather than the optimal solution was reached. It is suggested that the proposed method enables accurate and quick robotic navigation to find heart components.

Multi-Task Multi-Objective Evolutionary Search Based on Deep Reinforcement Learning for Multi-Objective Vehicle Routing Problems with Time Windows

The vehicle routing problem with time windows (VRPTW) is a widely studied combinatorial optimization problem in supply chains and logistics within the last decade. Recent research has explored the potential of deep reinforcement learning (DRL) as a promising solution for the VRPTW. However, the challenge of addressing the VRPTW with many conflicting objectives (MOVRPTW) still remains for DRL. The MOVRPTW considers five conflicting objectives simultaneously: minimizing the number of vehicles required, the total travel distance, the travel time of the longest route, the total waiting time for early arrivals, and the total delay time for late arrivals. To tackle the MOVRPTW, this study introduces the MTMO/DRP-AT, a multi-task multi-objective evolutionary search algorithm, by making full use of both DRL and the multitasking mechanism. In the MTMO/DRL-AT, a two-objective MOVRPTW is constructed as an assisted task, with the objectives being to minimize the total travel distance and the travel time of the longest route. Both the main task and the assisted task are simultaneously solved in a multitasking scenario. Each task is decomposed into scalar optimization subproblems, which are then solved by an attention model trained using DRL. The outputs of these trained models serve as the initial solutions for the MTMO/DRL-AT. Subsequently, the proposed algorithm incorporates knowledge transfer and multiple local search operators to further enhance the quality of these promising solutions. The simulation results on real-world benchmarks highlight the superior performance of the MTMO/DRL-AT compared to several other algorithms in solving the MOVRPTW.

A local space transfer learning-based parallel Bayesian optimization with its application

The optimization of process parameters in polyolefin production can bring significant economic benefits to the factory. However, due to small data sets, high costs associated with parameter verification cycles, and difficulty in establishing an optimization model, the optimization process is often restricted. To address this issue, we propose using a transfer learning Bayesian optimization strategy to improve the efficiency of parameter optimization while minimizing resource consumption. Specifically, we leverage Gaussian process (GP) regression models to establish an integrated model that incorporates both source and target grade production task data. We then measure the similarity weights of each model by comparing their predicted trends, and utilize these weights to accelerate the solution of optimal process parameters for producing target polyolefin grades. In order to enhance the accuracy of our approach, we acknowledge that measuring similarity in a global search space may not effectively capture local similarity characteristics. Therefore, we propose a novel method for transfer learning optimization that operates within a local space (LSTL-PBO). This method employs partial data acquired through random sampling from the target task data and utilizes Bayesian optimization techniques for model establishment. By focusing on a local search space, we aim to better discern and leverage the inherent similarities between source tasks and the target task. Additionally, we incorporate a parallel concept into our method to address multiple local search spaces simultaneously. By doing so, we can explore different regions of the parameter space in parallel, thereby increasing the chances of finding optimal process parameters. This localized approach allows us to improve the precision and effectiveness of our optimization process. The performance of our method is validated through experiments on benchmark problems, and we discuss the sensitivity of its hyperparameters. The results show that our proposed method can significantly improve the efficiency of process parameter optimization, reduce the dependence on source tasks, and enhance the method's robustness. This has great potential for optimizing processes in industrial environments.

Differential evolution based on strategy adaptation and deep reinforcement learning for multimodal optimization problems

Multimodal optimization problems (MMOPs) include multiple optima, which are common in practical fields. However, the success of solving MMOPs requires the algorithm to have both exploration and exploitation performance. How to select a reasonable search strategy is a difficult problem facing MMOPs. In this research, a differential evolution based on strategy adaptation and deep reinforcement learning, termed SA-DQN-DE, is proposed to select mutation strategies reasonably and search the optima effectively, which mainly includes three aspects. First, strategy adaptation, which calculates selection probability based on the feedback of different mutation operators in previous evolution, and assigns mutation operations to each individual to provide guidance for the next stage of evolution is developed. Secondly, a new historical individual preservation method is designed to improve search efficiency. Thirdly, deep reinforcement learning is applied to select multiple local search operators to refine the accuracy of potential optimal solutions. The performance of SA-DQN-DE is tested on publicly acknowledged CEC2013 benchmark MMOP set. The experimental results demonstrate that the proposed SA-DQN-DE has competitive performance compared with some of its peer multimodal optimization algorithms.

Fleet scheduling for electric towing of aircraft under limited airport energy capacity

Taxiing aircraft using electric vehicles is seen as an effective solution to meet aviation targets of climate neutrality. However, making the transition to electric taxiing operations is expected to significantly increase the electricity demand at airports. In this paper we propose a mixed-integer linear program to schedule electric vehicles for aircraft towing and battery charging, while considering a limit for the supply of energy. The objective of the schedule is to maximize emissions savings. For computational tractability, we develop an Adaptive Large Neighbourhood Search which makes use of multiple local search heuristics to identify scheduling solutions. For daily scheduling with a small fleet size, the developed heuristic achieves solutions with an average 4% gap to the best linear programming solution. The results show that charging the vehicles during daytime is essential to maximize saved emissions: removing charging opportunities for a few hours during the day reduces the performance by an average of 6.4%. In addition, it is found that fast charging leads to low vehicle downtime, unless the battery size exceeds 750kWh, when charging rates over 150kW become unnecessary. Overall, our model provides support for infrastructure planning of airports during the transition to aircraft electric taxiing.

A hybrid genetic algorithm for the Hamiltonian p‐median problem

AbstractThe Hamiltonian p‐median problem consists of finding p( is given) non‐intersecting Hamiltonian cycles in a complete edge‐weighted graph such that each cycle visits at least three vertices and each vertex belongs to exactly one cycle, while minimizing the total cost of pcycles. In this work, we present an effective and scalable hybrid genetic algorithm to solve this computationally challenging problem. The algorithm combines an edge‐assembly crossover to generate promising offspring solutions from high‐quality parents, and a multiple neighborhood local search to improve each offspring solution. To promote population diversity, the algorithm applies a mutation operator to the offspring solutions and a quality‐and‐distance update strategy to manage the population. We compare the method to the best reference algorithms in the literature based on three sets of 145 popular benchmark instances (with up to 318 vertices), and report improved best upper bounds for eight instances. To evaluate the scalability of the method, we perform experiments on a new set of 70 large instances (with up to 1060 vertices). We examine the contributions of key components of the algorithm.

Solving Flexible Shop Scheduling Problems Based on Improved NSGA-II

This paper proposes an improved non-dominated sorting genetic algorithm-II (NSGA-II) by incorporating local search techniques to address the issues of slow convergence speed and susceptibility to local optima in traditional NSGA. Multiple local search techniques are combined with NSGA-II to enhance the convergence speed and solution quality. In each generation, the algorithm selects key operations on the critical path for local search and applies the principles of minimizing costs and minimizing energy consumption to guide the local search on individuals. The optimization goals consist of reducing the longest completion time, minimizing expenses, and lowering energy usage. Experimental results on multiple instances of flexible job shop scheduling problems demonstrate that this algorithm exhibits good performance and effectiveness.

Application of surrogate-assisted global optimization algorithm with dimension-reduction in power optimization of floating offshore wind farm

The wake in a large-scale Floating Offshore Wind Farm (FOWF) can reduce the wind speed in downstream areas, thereby affecting and reducing the power production of FOWF. To maximize the power production of the FOWF, it is necessary to solve the power optimization problem and obtain the optimal control actions for the coordinated operation of wind turbines. Due to the complexity of the optimization problem, there remains difficulty in applying swarm intelligence algorithms, and thus we propose a dimensionality reduction-based surrogate-assisted framework. Firstly, a low-dimensional surrogate model is constructed using data generated by swarm intelligence algorithms and dimensionality reduction algorithms during the optimization process. Secondly, based on the low-dimensional surrogate model, multi-subswarms pre-screening is carried out to filter out poor solutions in the population and reduce the time consumption of calculating the objective function. Thirdly, the trust region range is determined and a local surrogate model is constructed for multiple local searches using the results of the multi-subswarms pre-screening, and the elite individuals obtained further guide the population individuals to search for optimization, improving the optimization efficiency of the algorithm. Finally, the proposed method is simulated and verified in a FOWF with 64 wind turbines. Under the wind direction where the wake effect is severe, the steady-state power production can be increased by more than 6%. Compared with conventional swarm intelligence algorithms and numerical solution methods, the proposed method produces more power while having smaller time cost under different wind conditions.

Redistricting optimization with recombination: A local search case study

In the U.S., states redraw electoral district boundaries every ten years. Given that redistricting affects political representation at both the state and national levels, it is crucial to prevent the manipulation of district boundaries for political gain. Optimization methods can be valuable tools for promoting transparency and fairness in redistricting. Here we examine a novel local search approach for redistricting that transitions between feasible solutions using Recombination (a recently introduced spanning tree iteration). We compare the performance of multiple local search heuristics using both Recombination and more traditional Flip iterations by optimizing congressional plans for Illinois, Missouri, and Tennessee with respect to several common fairness objectives. We evaluate which heuristic produces the best objective value within limited time periods and generate collections of optimized plans. The Recombination heuristics produced excellent objective values, often far superior to the Flip heuristics; they also maintained more compact district shapes when the objective was not compactness. However, the Flip heuristics converged to a local optimum more quickly and occasionally achieved better solutions than the ReCom heuristics within short time periods. Hence, while the use of Recombination within local search frequently improves solution quality, there are some scenarios for which Flip may be preferable.

Dual-Dimensional Manufacturing Service Collaboration Optimization Toward Industrial Internet Platforms

An Industrial Internet platform is acknowledged to be a requisite promoter for smart manufacturing, enabling physical manufacturing resources to be virtualized and permitting resources to collaborate in the form of services. As a central function of the platform, manufacturing service collaboration optimization is dedicated to establishing high-quality service collaboration solutions for manufacturing tasks. Such optimization is inseparable from the functional and amount requirements of a task, which must be satisfied when orchestrating services. However, existing manufacturing service collaboration optimization methods mainly focus on horizontal collaboration among services for functional demands and rarely consider vertical collaboration to cover the needed amounts. To address this gap, this paper proposes a dual-dimensional service collaboration methodology that combines functional and amount collaboration. First, a multi-granularity manufacturing service modeling method is presented to describe services. On this basis, a dual-dimensional manufacturing service collaboration optimization (DMSCO) model is formulated. In the vertical dimension, multiple functionally equivalent services form a service cluster to fulfill a subtask; in the horizontal dimension, complementary service clusters collaborate for the entire task. Service selection and amount distribution to the selected services are critical issues in the model. To solve the problem, a multi-objective memetic algorithm with multiple local search operators is tailored. The algorithm embeds a competition mechanism to dynamically adjust the selection probabilities of the local search operators. The experimental results demonstrate the superiority of the algorithm in terms of convergence, solution quality, and comprehensive metrics, in comparison with commonly used algorithms.

An Iterated Local Search Methodology for the Qubit Mapping Problem

The qubit mapping approach serves to transform a quantum logical circuit (LC) into a physical one that satisfies the connectivity constraints imposed by the noisy intermediate-scale quantum (NISQ) devices. The quality of the physical circuit generated by a mapping approach depends largely on the initial mapping, which specifies the correspondence between the qubits in the LC and the qubits on the NISQ device. There are a total of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$n!$ </tex-math></inline-formula> different initial mappings for a qubit mapping problem with <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$n$ </tex-math></inline-formula> qubits, and among them, there is at least one initial mapping corresponding to the smallest physical circuit that this mapping approach can output. Finding such an initial mapping is very important for reliable computations on the NISQ device. To this end, we propose an iterated local search framework as well as a heuristic circuit mapper. In this framework, we perform multiple local searches on the space of initial mappings, and during each local search, several promising neighborhoods of the current initial mapping are generated and evaluated by invoking the circuit mapper in a forward or a backward manner. This framework provides a way for the qubit mapping approach to find the best physical circuit that it can produce, allowing it to trade time for circuit quality, which is necessary in the NISQ era. The experimental results demonstrate the stability, scalability, and effectiveness of this approach in reducing the number of additional gates. Moreover, although this approach is a multipass circuit mapping process, it can generate a good-quality physical circuit within half an hour, even for the circuit with more than 10 000 gates.

Hierarchical Collaborated Fireworks Algorithm

The fireworks algorithm (FWA) has achieved significant global optimization ability by organizing multiple simultaneous local searches. By dynamically decomposing the target problem and handling each one with a sub-population, it has presented distinct property and applicability compared with traditional evolutionary algorithms. In this paper, we extend the theoretical model of fireworks algorithm based on search space partition to obtain a hierarchical collaboration model. It maintains both multiple local fireworks for local exploitation and one global firework for overall population distribution control. The implemented hierarchical collaborated fireworks algorithm is able to combine the advantages of both classic evolutionary algorithms and fireworks algorithms. Several experiments are provided for in-depth analysis and discussion on the proposed algorithm. The effectiveness of proposed strategy is demonstrated on the benchmark test suite from CEC 2020. Experimental results validate that the hierarchical collaborated fireworks algorithm outperforms former fireworks algorithms significantly and achieves similar results compared with state-of-the-art evolutionary algorithms.

Efficient multiobjective optimization for an AGV energy-efficient scheduling problem with release time

In recent years, green manufacturing has attracted wide attention from researchers. However, the energy efficiency problem in matrix manufacturing workshops is still a blank area. This paper considers a novel automatic guided vehicle (AGV) energy-efficient scheduling problem with release time (AGVEESR) to optimize the three objectives of energy consumption, number of AGVs used and customer satisfaction simultaneously. Considering the development of the AGVEESR, we extract problem-specific knowledge, establish a multiobjective mathematical model, and design a hybrid constructive heuristic. Due to the complexity of the problem, we propose an efficient multiobjective greedy algorithm (MOGA) with effective strategies such as new population initialization, greedy operation, and self-adaptive multiple neighbourhood local search. Meanwhile, an ideal-point-based construction operator in the greedy operation phase is presented to lower the computational complexity. Simulation results show that the proposed MOGA has a tremendously superior performance to the five state-of-the-art algorithms in solving the problem considered.

Particle Swarm Optimization Based on Fitness Considering Distance Information for Multiple Local Optimal Solutions Search

Particle Swarm Optimization Based on Fitness Considering Distance Information for Multiple Local Optimal Solutions Search

Multi-objective evolutionary algorithm based on multiple neighborhoods local search for multi-objective distributed hybrid flow shop scheduling problem

In order to be competitive in today’s rapidly changing business world, enterprises have transformed a centralized to a decentralized structure in many areas of decision. It brings a critical problem that is how to schedule the production resources efficiently among these decentralized production centers. This paper studies a multi-objective distributed hybrid flow shop scheduling problem (MDHFSP) with the objectives of minimizing makespan, total weighted earliness and tardiness, and total workload. In the MDHFSP, a set of jobs have to be assigned to several factories, and each factory contains a hybrid flow shop scheduling problem with several parallel machines in each stage. A multi-objective evolutionary algorithm based on multiple neighborhoods local search (MOEA-LS) is proposed to solve the MDHFSP. In the initialization phase, a weighting mechanism is used to decide which position is the best one for each job when constructing a new sequence. Several multiple neighborhoods local search operators based on the three objectives are designed to generate offsprings. Some worse neighboring solutions are replaced by the solutions in the achieve set with a simulated annealing probability. In order to avoid trapping into local optimum, an adaptive weight updating mechanism is utilized when the achieve set has no change. The comprehensive comparison with other classic multi-objective optimization algorithms shows the proposed algorithm is very efficient for the MDHFSP.

Pattern mining‐based pruning strategies in stochastic local searches for scheduling problems

AbstractScheduling problems are a subclass of combinatorial problems consisting of a set of tasks/activities/jobs to be processed by a set of resources usually to minimize a time criterion. Some optimization methods used to solve these problems are hybridized with knowledge discovery techniques to extract information during the optimization process and enhance it. However, most of these hybrid techniques are custom‐designed and lack generalization. In this paper, a module for knowledge extraction in Stochastic Local Searches is designed, aiming to be problem independent and plugged into optimization methods that relies on multiple Stochastic Local Search replications. The objective is to prune parts of the search space for which the exploration is likely to lead to poor solutions. This is performed through the extraction of high‐quality patterns occurring in locally optimal solutions. Benchmarked on two well‐known scheduling problems, the Job‐shop Problem and the Resource Constrained Project Scheduling Problem, the results show both a speed up in the convergence and the reaching of better local optima solutions.

Local search-based metaheuristics for the robust distributed permutation flowshop problem

Distributed Permutation Flowshop Scheduling Problem (DPFSP) has become a hot issue in recent years. However, DPFSP with uncertain processing times (DPFSP_UPT) has not been addressed so far although real manufacturing systems often encounter various uncertain factors which make processing times uncertain. In this paper, two local-search based metaheuristics, i.e., an Iterated Greedy algorithm (sIG) and an Iterated Local Search algorithm (sILS), are presented to solve the DPFSP_UPT with makespan criterion. A robust model of expect-risk rule is used to describe the DPFSP_UPT. A valid heuristic based on the well-known NEH heuristic is proposed to initialize sIG and sILS. An acceleration algorithm is adapted to save computational efforts. A destruction procedure with dynamic sizes is presented to enhance the exploration capability for sIG. A feature with the characteristic that small changes in the solution can preserve the properties of good solutions in the iterative process is applied by sILS. Both sIG and sILS use multiple local search methods to produce promising solutions by exploiting diverse search areas. Extensive experiments show that the proposed sIG and sILS perform significantly better than the five competing algorithms adapted in the literature, but there is no significant difference between sIG and sILS and each has its own advantages for different instances.

Niching Migratory Multi-Swarm Optimization for Multiple Local Minimum Search: Towards Constant population and Concurrent Search by Physical Robots

Niching Migratory Multi-Swarm Optimization for Multiple Local Minimum Search: Towards Constant population and Concurrent Search by Physical Robots

A novel local search method for LSGO with golden ratio and dynamic search step

Depending on the developing technology, large-scale problems have emerged in many areas such as business, science, and engineering. Therefore, large-scale optimization problems and solution techniques have become an important research field. One of the most effective methods used in this research field is memetic algorithm which is the combination of evolutionary algorithms and local search methods. The local search method is an important part that greatly affects the memetic algorithm’s performance. In this paper, a novel local search method which can be used in memetic algorithms is proposed. This local search method is named as golden ratio guided local search with dynamic step size (GRGLS). To evaluate the performance of proposed local search method, two different performance evaluations were performed. In the first evaluation, memetic success history-based adaptive differential evolution with linear population size reduction and semi-parameter adaptation (MLSHADE-SPA) was chosen as the main framework and comparison is made between three local search methods which are GRGLS, multiple trajectory search local search (MTS-LS1) and modified multiple trajectory search. In the second evaluation, the improved MLSHADE-SPA (IMLSHADE-SPA) framework which is a combination of MLSHADE-SPA framework and proposed local search method (GRGLS) was compared with some recently proposed nine algorithms. Both of the experiments were performed using CEC’2013 benchmark set designed for large-scale global optimization. In general terms, the proposed method achieves good results in all functions, but it performs superior on overlapping and non-separable functions.

Hyper-heuristic local search for combinatorial optimisation problems

Local search algorithms have been successfully used for many combinatorial optimisation problems. The choice of the most suitable local search algorithm is, however, a challenging task as their performance is highly dependent on the problem characteristic. In addition, most of these algorithms require users to select appropriate internal neighbourhood structures to obtain desirable performance. No single local search algorithm can consistently perform well with a fixed setting, for different types of problems or even different instances of the same problem. To address this issue, we propose a hyper-heuristic framework which incorporates multiple local search algorithms and a pool of neighbourhood structures. This framework is novel in three respects. Firstly, a two-stage hyper-heuristic structure is designed to control the selection of a local search algorithm and its internal operators. Secondly, we propose an adaptive ranking mechanism to choose the most appropriate neighbourhood structures for the current local search algorithm. The proposed mechanism uses the entropy to evaluates the contribution of the local search in terms of quality and diversity. It adaptively adjusts the pool of candidate neighbourhood structures. Thirdly, we use a population of solutions within the proposed framework to effectively navigate different areas in the solutions search space and share solutions with local search algorithms. To ensure different solutions is allocated in different regions of the search space, we propose a distance-based strategy for population updating process that allowing solutions to share local search algorithms. We have evaluated the performance of the proposed framework using two challenging optimisation problems: Multi-Capacity Bin Packing benchmark instances and Google Machine Reassignment benchmark instances. The results show the effectiveness of the proposed framework, which outperformed state-of-the-art algorithms on several problem instances.