Search

Optimization methods are widely used in various fields to find the best possible solution to a given problem through the minimization or maximization of an objective function while adhering to specific constraints. In this paper, we present a new optimization algorithm, called WOADD, which was designed to handle the challenges of constrained optimization problems that involve dependent data. Unlike traditional algorithms that struggle with data dependencies and valid range constraints, WOADD uses a unique normalization process and a dynamic updating mechanism that accurately considers the interdependencies among features. By calculating a scaling parameter to navigate within feasible regions, WOADD adjusts its search strategy to ensure the preservation of data dependencies and adherence to constraints, ultimately leading to more efficient and precise optimization outcomes. Our extensive experimental analysis, which compared WOADD against other swarm optimization methods using a suite of benchmark functions, demonstrates its superior performance in terms of faster convergence rates, improved solution quality, and enhanced determinism in outcomes. [ABSTRACT FROM AUTHOR]

This paper proposes a PID controller optimized by a reformative artificial bee colony algorithm (RABC-PID) for the radar servo system (RSS). The RABC algorithm is an enhancement of the artificial bee colony (ABC) algorithm by introducing the best-positioned food source and modifying the food source probability. The RABC algorithm is validated by simulation with six benchmark functions, and the results show that the RABC algorithm is superior to the other variants of the ABC algorithm in terms of convergence speed and accuracy. The RABC-PID controller is then used for the RSS. The RSS is presented to illustrate the application of the RABC-PID controller. The simulation results, which are also compared to PID optimized by particle swarm optimization, differential evolution, and genetic algorithm (PSO-PID, DE-PID, and GA-PID) respectively, are shown to illustrate the effectiveness and robustness of the RABC-PID controller. [ABSTRACT FROM AUTHOR]

Recently a continuous description of particle swarm optimization (PSO) based on a system of stochastic differential equations was proposed by Grassi and Pareschi in [Math. Models Methods Appl. Sci., 31 (2021), pp. 1625—1657] where the authors formally showed the link between PSO and the consensus-based optimization (CBO) through the zero-inertia limit. This paper is devoted to solving this theoretical open problem proposed in [S. Grassi and L. Pareschi, Math. Methods Appl. Sci., 31 (2021), pp. 1625—1657] by providing a rigorous derivation of CBO from PSO through the limit of zero inertia, and a quantified convergence rate is obtained as well. The proofs are based on a probabilistic approach by investigating both the weak and strong convergence of the corresponding stochastic differential equations of Mckean type in the continuous path space and the results are illustrated with some numerical examples. [ABSTRACT FROM AUTHOR]

The vendor managed inventory (VMI) is the most commonly known marketing and distribution methodology among the service provider and the retailer in a supply chain environment. The VMI framework's critical feature is to satisfy the demand of the products or items by the supplier rapidly generated from the retailers with the number of orders (order quantities). Due to the business's changing conditions, necessities of the particular product, production cycle, and manufacturing expenses, the VMI system's demand and order quantity are highly uncertain. Therefore, the total cost of a VMI system possesses higher-order uncertainties for a real-world scenario. This paper proposes a novel interval type-2 fuzzy vendor managed inventory (IT2FVMI) system, in which demand and order quantity are represented by the interval type-2 fuzzy numbers. The proposed IT2FVMI model aims to minimize the overall cost for the single vendor-retailer business, merchandise of multi-product, and a centralized warehouse of retailers' inventory managed by the vendor. Since the proposed model is an NP-hard, a particle swarm optimization (PSO) based solution approach is developed for solving it appropriately. Moreover, we have also formulated the classical/crisp VMI model and type-1 fuzzy vendor managed inventory (T1FVMI) model via considering demand and order quantity as the deterministic and the type-1 fuzzy numbers, respectively. The proposed solution technique is capable of solving both crisp VMI and T1FVMI problems. An appropriate real-world application is considered to conduct experimental simulations for the five test problems that displayed the various situations, and the minimum costs of the models are obtained. All the three models are thoroughly analyzed, and from comparison, it is demonstrated that the proposed IT2FVMI model outperforms both T1FVMI and crisp VMI models by providing a more trustworthy solution in terms of minimum cost, statistical analysis, and significantly faster convergence. [ABSTRACT FROM AUTHOR]

Particle swarm optimization (PSO) is a well-known optimization algorithm that shows good performance in solving different optimization problems. However, PSO usually suffers from slow convergence. In this article, a reinforcement learning strategy is developed to enhance PSO in convergence by replacing the uniformly distributed random number in the updating function with a random number generated from a selected normal distribution. In the proposed method, the mean and standard deviation of the normal distribution are estimated from the current state of each individual through a policy net. The historic behaviour of the swarm group is used to update the policy net and guide the selection of parameters of the normal distribution. The proposed method is integrated into the original PSO and a state-of-the-art PSO, called the self-adaptive dynamic multi-swarm PSO (sDMS-PSO), and tested with numerical functions and engineering problems. The test results show that the convergence rate of PSO methods can be improved with the proposed reinforcement learning strategy. [ABSTRACT FROM AUTHOR]