International Journal of Fuzzy Systems最新文献

Fish diseases are among the major limiting factors to increase global aquaculture production. They lead to increased fish mortality, low breeding and growth rates, and low meat quality. The success of aquaculture is heavily dependent on the timely identification of disease. Therefore, we propose a fuzzy U-Net model to automatically identify fish disease from underwater images. U-Net is enhanced with multi-head channel and spatial attention and used to segment infected fish regions from fish disease images. Color pixel intensity features are then extracted from the localized regions, a form of guided feature extraction. Fuzzy C-means clustering is then used to find the cluster centroids and data distribution within the clusters, for the design of fuzzy membership functions. Moreover, the number of clusters are determined by silhouette score. Adaptive neuro-fuzzy inference system (ANFIS) is then trained, tested, and cross-validated for fish disease identification. The model parameters are optimized using particle swarm optimization (PSO) algorithm and compared with gradient-based methods. For image segmentation, the enhanced U-Net achieved a mean intersection over union (IoU) of 86.29%, mean pixel accuracy of 90.94%, mean precision of 93.58%, and mean recall value of 89.94% on 42 test images. Subsequently, ANFIS with PSO achieved overall superior performance on fish disease identification over gradient-based methods, with accuracy of 99.31%, precision of 99.00%, recall of 99.00%, and F1-score of 99.00%. The high-performance results of the optimized ANFIS confirm the robustness and efficacy of the proposed method to automatically identify fish diseases in aquaculture.

Safety life analysis can provide guidance for safety service and maintenance plan of structure. To efficiently analyze the structural safety life under required failure possibility in the presence of fuzzy uncertainty, this paper proposes a sequential Kriging (SK)-based fuzzy simulation (FS) combined with dichotomy (SK-FS-D) method. Firstly, based on monotonic relationship of time-dependent failure possibility (TDFP) and service time, the SK-FS-D uses dichotomy method to search the safety life. Secondly, the SK-FS-D proposes a strategy of sequentially updating Kriging surrogate model in the candidate sample pool (CSP) of fuzzy simulation (FS) to estimate the TDFP corresponding to each possible life searched by dichotomy, and the next dichotomy interval is determined by TDFP estimated by the convergent Kriging model. This strategy can improve the efficiency of SK-FS-D by avoiding training Kriging model at the life not visited by dichotomy and extend the engineering applicability of SK-FS-D by inheriting the advantages of FS method. Moreover, the CSP reduction strategy is further adopted to improve the computational efficiency of TDFP according to some correct information provided by the convergent Kriging model and the property of fuzzy design point. Finally, one numerical example and three engineering examples are introduced to verify the superior performance of the proposed SK-FS-D over the existing methods.

As an essential tool for processing fuzzy or chaotic information, the main feature of the first-order Takagi–Sugeno (T–S) neuro-fuzzy model is utilizing a set of IF-THEN fuzzy rules to represent non-linear systems, showcasing commendable non-linear approximation ability and significant interpretability. However, the coexistence of linear rules and the affiliation function of fuzzy sets makes the integer-order gradient descent method (IOGDM), commonly used in training the first-order T–S neuro-fuzzy model, fail to accurately capture the intricate relationships among weights, resulting in the error function struggling to converge rapidly to low values. To enhance the convergence speed and training accuracy of the first-order T–S neuro-fuzzy model during the training process, a fractional-order gradient descent method (FOGDM) is proposed to update the fuzzy rule parameters and neural network weights of the model in this paper. By subdividing the gradient into fractional orders, FOGDM exhibits heightened flexibility in gradient adjustments, thus better capturing the complex non-linear relationships among parameters during the optimization process. The weak and strong convergence of the proposed approach is meticulously demonstrated in this paper, ensuring that the weight of error functions converges to a constant value and that the gradient of the error functions tends toward zero, respectively. Simulation results analysis indicates that, compared to IOGDM, FOGDM exhibits faster convergence speed and more significant generalization capabilities.

This work researches the issue of adaptive fault-tolerant fuzzy tracking control for a class of nonlinear systems in strict-feedback form with quantized inputs. The fuzzy logic systems are utilized to approximate unknown functions, and a fuzzy state observer is built to estimate the unavailable states. Meanwhile, an improved hysteresis quantizer is introduced to achieve the quantized inputs for saving communication resources. To improve the approximation capacities of fuzzy logic systems, the compensated tracking errors and the prediction errors are used to construct the adaptive laws parameters. Furthermore, a composite adaptive fault-tolerant fuzzy control strategy is developed, which can guarantee proper operations of the systems when encountering actuator faults, and overcome the issue of “explosion of complexity” in the backstepping approach. It is strictly demonstrated that the system output can follow a desired signal within a small error zone and all signals of the closed-loop system are bounded. Finally, the simulation results are given to confirm the validity of the presented control strategy.

Rough fuzzy K-means (RFKM) decomposes data into clusters using partial memberships by underlying structure of incomplete information, which emphasizes the uncertainty of objects located in cluster boundary. In this scheme, the settings of cluster boundary merely depend on subjective judgment of perceptual experience. When confronted with the data exhibiting heavily overlap and imbalance, the boundary regions obtained by existing empirical schemes vary greatly accompanied by skewing of cluster center, which exerts considerable influence on the accuracy and stability of RFKM. This paper seeks to analyze and address this deficiency and then proposes an improved rough fuzzy K-means clustering based on parametric decision-theoretic shadowed set (RFKM-DTSS). Three-way approximation is implemented by incorporating a novel fuzzy entropy into the decision-theoretic shadowed set, which rationalizes cluster boundary through minimizing fuzzy entropy loss. Under the secondary adjustment method and improved update strategy of cluster center, the proposed RFKM-DTSS is thus featured by a powerful processing ability on class overlap and imbalance commonly seen in scenarios, such as fault detection and medical diagnosis with unclear decision boundaries. The effectiveness and robustness of the RFKM-DTSS are verified by the results of comparative experiments, demonstrating the superiority of the proposed algorithm.

In real-world scenarios, datasets often lack full supervision due to the high cost associated with acquiring decision labels. Completing datasets by filling in missing labels is essential for preserving the valuable feature information of individual samples. Furthermore, in the era of big data, datasets tend to exhibit high dimensionality, which adds complexity to subsequent data processing. In this study, a new semi-supervised feature selection technique is introduced. Firstly, a fully supervised dataset is created by utilizing a local density decision-labeling algorithm to fill in missing decision labels within the semi-supervised dataset. Next, a fuzzy dependency-based feature selection approach is presented to find and keep the most pertinent characteristics for the finished datasets. Finally, the effectiveness and reliability of our proposed method are validated through a series of rigorous experiments.

Faced with rapidly rising energy demand in industrialised societies and widespread global concern, countries are actively promoting the transition from conventional to renewable energy systems. The goal is to invest in renewable energy in the most efficient way to meet rising energy demand and reduce the challenges posed by climate change. However, decision makers must carefully weigh various factors when selecting the most appropriate renewable energy investment projects. This paper presents a novel method for Multi-Attribute Decision Making(MADM) that uses the Bipolar Complex Fuzzy(BCF) to convey the vagueness and uncertainty of decision makers, so that the result obtained better reflects the actual scenario and the subjective biases of decision makers. We defined BCF Einstein Weighted Averaging (BCFEWA) operator and BCF Einstein Ordered Weighted Averaging (BCFEOWA) operator to aggregate evaluation information. Then we discussed some properties of the proposed aggregation operators. Additionally, we present an integrated MADM technique grounded in the BCF framework that combines the CRiteria Importance Through Intercriteria Correlation (CRITIC) and ELECTRE III methods. Specifically, the CRITIC method determines attribute weights, and the ELECTRE III method ranking the alternatives to determine the best renewable energy investment projects. After analysing the results and comparisons, it can be inferred that the suggested methodology offers an effective evaluation process.

The remote sensing satellite observation process involves multiple stakeholders and significant costs, so selecting an appropriate observation scheme and reaching an agreement on the chosen scheme among the evaluators/stakeholders is essential. From this perspective, the observation scheme selection problem can be viewed as a large-scale group decision-making (LSGDM) problem, challenging due to its complex group composition and the high consensus level required. Accordingly, this paper investigates an adaptive bi-directional consensus model that incorporates the evolution of social influence to address the LSGDM problem. Firstly, the dual-attribute affinity propagation algorithm is employed to divide the large-group into manageable subgroups. Secondly, the social influence evolution model is established, where evaluators’ social influences are determined by considering their opinion similarity and trust level, and subgroups’ social influences are updated by measuring their decision risk. Thirdly, the bi-directional feedback mechanism is designed to adaptively generate adjustment strategies corresponding to different scenarios based on the evolution model. Finally, an observation scheme selection case is analyzed using the proposal to demonstrate its practicality. During the process of remote sensing satellite observation, the selection of an appropriate observation scheme can optimize the utilization of existing satellite resources and ensure the quality of satellite observation services, thereby better meeting the demands of diverse application areas such as environmental monitoring, disaster management, and urban planning.

In a volatile stock market, an investor’s long-term goal involves determining the most effective buying, selling strategies, and money management techniques in order to maximize profits. This paper introduces a multi-agent trading system to achieve this goal, termed QF-FRL, based on quantum finance and fuzzy reinforcement learning (QF-FRL). The system comprises two agents: (1) The trading agent, constructed using the Deep Deterministic Policy Gradient (DDPG) and Twin Delayed Deep Deterministic Policy Gradient (TD3). This agent employs a Denoising Auto Encoder (DAE) to extract stock representations from historical time series data. The trading agent initially employed the DDPG model, which was subsequently supplanted by the TD3 model. It integrates traditional financial technology indicators, like moving averages, with modern deep reinforcement learning technology to generate buying and selling signals for determining the optimal strategy. (2) The risk control agent, founded on quantum finance and an adaptive network-based fuzzy inference system. This agent merges the QPL indicator with a fuzzy risk control method to ascertain transaction amounts. Furthermore, a genetic algorithm is utilized to optimize the parameters of the fuzzy system, aiming to enhance profits and ensure accuracy in transactions at specific amounts. The experiments in this study involved selecting nine stocks and testing them against seven competing quantitative trading models. Upon comparing the profit rate, trading frequency, Sharpe ratio, and average return of each stock, eight stocks within the QF-FRL system achieved the highest returns and a greater number of transactions. Additionally, the QF-FRL system has also attained the highest average return and the second highest average Sharpe ratio. The results indicate that QF-FRL outperforms competing models, yielding higher profits and being particularly suitable for long-term investment. Moreover, it exhibits more favorable risk-adjusted returns and a notable degree of robustness.

This paper is devoted to the observer-based adaptive robust control for fractional-order Takagi–Sugeno (T-S) fuzzy systems with input uncertainties and output disturbances. By combining system states and output perturbations as new state variables, an augmented fuzzy system whose state variables are unknown is built. Furthermore, an observer is devised to simultaneously estimate unmeasurable system states together with unknown external disturbances. Two stability theorems are derived to prove the asymptotic stability of the error system based on linear matrix inequalities and Lyapunov stability theory. Finally, simulation results are provided to demonstrate the effectiveness of the designed method.