Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Aifan Ling)

    We developed a sailboat pricing model that takes into account the different impacts of different boat characteristics on pricing, as well as the important role of brokers and pricing strategies in the boat selling process. The main points of the methodology are: first, we performed feature screening to filter out the main factors affecting pricing, then we constructed a BP neural network regression model to analyze the impact of various factors on pricing, and finally, we used an ant colony algorithm to optimize the parameters of the BP neural network regression model to produce the results. Prior to this, we preprocessed the collected data and introduced dummy variables to improve the quality of the data as the data substituted into the model is highly susceptible to noise, missing values and inconsistent data. Next, we apply the model to evaluate the weights of each factor affecting pricing between single-hull and double-hull vessels. Immediately after we substitute different ship types from different regions into the model, the results show that the weights predicted by our developed model are closer to the actual values than those predicted by the BP neural network model, the random forest model, the XGBoost model, and the LightGBM model, which proves the value of the model when it is applied to the pricing of sailing vessels. Finally we use the model to predict the price of single hull and double hull vessels in different regions using box plots to show the average price and the results show that Caribbean is the cheapest among Catamarans and the most expensive among Monohulled Sailboats.And we show that the stability of the model is greatly improved with the introduction of dummy variables, making the model highly adaptable.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Shijing Si)

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Aifan Ling)

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

    Sound field control is widely used in various scenarios to create distinct acoustic environments within specific zones. Two novel low-order filter design methods for sound field control have been proposed by integrating finite difference penalties of L1-norm and L2-norm with the ACC-PM method. This approach enables the generation of low-order time-domain filters for the loudspeaker array, while accurately reproducing the target amplitude and phase distribution in a specific sound field. The alternating direction method of multipliers (ADMM) algorithm is employed to solve the  proposed models. Numerical experimental results demonstrate that, compared to four benchmark sound field control methods, the two proposed methods can effectively reduce the filter order.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

    This paper presents a novel hybrid target positioning methodology that integrates time difference of arrival (TDOA) measurements, kalman filter (KF) estimation, and particle swarm optimization (PSO) algorithm for enhanced maritime target tracking. The proposed framework is specifically designed for accurate localization of moving sea-surface targets using a swarm of unmanned aerial vehicles (UAVs). The methodology comprises three key components: first, a Chan-Taylor hybrid TDOA algorithm is developed to enhance both temporal resolution and system robustness compared to conventional Chan-based and Taylor-based TDOA approaches. Second, the system incorporates a KF-based predictive model that optimally fuses real-time positioning data with predicted target trajectories, thereby improving positioning precision. Third, the PSO algorithm is implemented to dynamically optimize UAV swarm configurations, maximizing the accuracy of the hybrid TDOA measurements. Extensive numerical simulations demonstrate the superior performance of the proposed method across multiple error metrics, confirming its effectiveness in maritime target tracking applications.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

    Data Envelopment Analysis (DEA) is a powerful method that evaluates performance efficiency through the use of multiple measures within this analytical framework. Among these measures, the non-radial graph model plays a prominent role and has garnered significant scholarly attention. Although the literature contains numerous studies examining individual DEA models and suggesting improvements, there remains a notable gap in comprehensive, systematic investigation of non-radial graph models within DEA, particularly concerning the relationships between measurement investigations, intrinsic properties of measures, and their varied applications. This paper addresses this gap by offering a systematic review of classical non-radial graph models in DEA, illustrated through two foundational frameworks: the Russell-type measure and the slack-variable approach. Moreover, with a focus on economic management, this paper showcases recent extensive applications of non-radial graph models across various fields. These include crucial roles in assessing global economic performance, analyzing regional economies, and evaluating efficiencies within financial markets. Through this thorough exploration, the paper not only enriches the theoretical foundation of non-radial graph models but also underscores their practical applications.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

    In this paper, we consider a class of multi-objective optimal control problems with a novel linear parameter varying (LPV) system. Different from the classic LPV systems, the scheduling parameters of the novel LPV systems are no longer preset but adjusted according to the output at sampling points. To solve the corresponding multi-objective optimal control problem, we first employ the control parameterization method to transform it into an optimization problem. We then propose the ICMA-MOEA/D algorithm to solve the transformation problem, which employs a segmented evolutionary approach that balances convergence and diversity by defining new DE operators at each stage and employing different aggregation functions. Numerical results show that compared with the original ICMA algorithm and NSGA II algorithm, the improved algorithm leads to a uniform population distribution and ultimately good tracking results.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

     This paper addresses a sparse initial source identification problem with linear diffusion-advection equations and a given final time. The problem's complexity arises from the initial source's structure, a linear combination of Dirac measures.  Moreover, the strong diffusion and smoothing effects of the linear diffusion-advection equation further render the problem exponentially ill-posed and non-smooth. To overcome these obstacles, we propose a two-stage numerical method. In the first stage, we reformulate the identification problem as a sparse optimal control problem, incorporating $L^1$ and $L^2$ regularization terms into the cost function to encourage sparsity and ensure well-posedness. To solve the sparse optimal control problem efficiently, augmented Lagrangian technique is extent to an inexact setting, where the sub-problem in each iteration is addressed using the semi-smooth Newton method. For the large-scale and ill-conditioned Newton system involved,  an efficient preconditioned conjugate gradient method is designed. Additionally, we demonstrate that with the appropriate selection of the penalty parameter, the proposed inexact augmented Lagrangian algorithm can achieve super-linear convergence rate. At the second stage, we refine the solution obtained at the first stage into a linear combination of Dirac measures. This process entails a location identification procedure and solving a small-scale least squares fitting problem. Lastly, three types of sparse initial source identification problems are considered and solved to demonstrate the feasibility and effectiveness of the proposed algorithm.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

    Many optimization problems modelled from actual problems have multiple global or local minimizers, which makes traditional optimization algorithms easily plunge into the local minimizer. The filled function algorithm is an effective method for such problems, which transforms the objective function into a filled function with special properties at the current local minimizer. By iteratively minimizing the objective and the filled function, the filled algorithm can gradually discover better local minimizers until the global minimizer is obtained. Hence, constructing new forms of filled function is a popular research direction, which can improve the performance of the filled function algorithm. A new parameter-free filled function with some advantages is constructed and the corresponding theoretical analyses are conducted in this paper. Firstly, its local minimizers are those of the objective function. Secondly, the current local minimizer of the objective function is the global maximizer of the proposed filled function. Based on the proposed filled function, a new global optimization algorithm without parameters to tune is given, which can avoid alternately solving the objective function and the filled function. The new algorithm is applied to solve several test functions with satisfactory experimental results, which show that the proposed algorithm is more efficient compared to the existing filled function algorithm.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Changjun Yu)

     To solve the issue of low efficiency in interpreting multidimensional unlabeled well logging data, an improved graph regularized non-negative matrix factorization model is proposed. This model takes into account the construction of basis matrix based on low-rank characteristics of the well logging data. Firstly, the rank of basis matrix is estimated by the low-rank matrix recovery model and weighted nuclear norm optimization algorithm. Secondly, the local features of the basis matrix are represented by a spectral clustering on the clean low-rank part. The local features and the optimized low-rank value are integrated to construct the basis matrix, which appropriately reflects the structure that is latent in the data. Finally, the optimized basis matrix is employed and visualized low dimensional non-negative features are extracted for well logging data within the graph regularized non-negative matrix factorization (GNMF) framework. Experimental comparisons on three well logging datasets and their combinations demonstrate the effectiveness of the proposed model, aiding in the efficient and automatic identification of the oil-bearing in reservoirs with unlabeled well logging data.

Special Issue:Some Optimization Methods in Artificial Intelligence, Finance, Control, Theory and Real-World Applications of Optimization

(Communicated by Lei Wang)

     In this study, we examine a linear time-invariant system with the influence of a static state feedback control mechanism, which takes the form of $Kx(t-\tau(||K||_0))$, where $K$ represents the gain matrix and $||K||_0$ denotes the number of nonzero entries of the matrix $K$. The parameter $\tau(||K||_0)$ corresponds to a varying delay, which arises due to the time required for the transmission of the system state information and the subsequent computation of the control input. Governed by the linear time-invariant system, we minimize prescribed conventional cost functions as $J^{0}(K)$ to obtain the optimal feedback matrices $K_1^{*}$. Numerous computational methods are available for the achievement of this objective. Nevertheless, it is noteworthy that the resulting $K_1^{*}$ matrix typically exhibit a dense structure. The primary objective of this paper is to minimize  the $l_0$ norm of the feedback matrix $K$ under conditions that satisfy the constraint $|J^{0}(K)-J^{0}(K_1^{*})|\leq \varepsilon$. The $l_0$ norm acts as a quantifiable measure for assessing the degree of sparsity within the feedback matrix. The sparsity of the gain matrix is approximated through the application of a piecewise quadratic approximation (PQA) of the $l_0$-norm of the feedback matrix. Subsequently, we proceed to formulate an iterative algorithm designed to address the transformed problem, accompanied by a thorough analysis of its convergence properties. Finally, we undertake a numerical experiment employing the proposed algorithm, aiming to illustrate its practical utility and efficacy in solving the problem at hand.