Highly accurate motion prediction is critical to facilitate safe navigation and autonomous control of a ship. In this paper, a math-data integrated prediction (MDIP) model of ship maneuvering motion is newly developed by combining with mathematical and data-driven modules. The math part is identified by employing variable-order hydrodynamic derivatives which are derived from Taylor expansion. Using extended Kalman filter, an optimal-order mathematical prediction model is obtained. The data-driven part dwells on prediction residuals which are approximated by a least square-support vector machine. Furthermore, integration mechanisms are devised to cohere mathematical and data-driven models as a whole, by deploying summation and neural network approximation, respectively. By comparisons, it is verified that not only order selection but also math-data integration can enhance ship motion prediction accuracy, for which various maneuvering tests are analyzed. The results demonstrate that the proposed MDIP model using math-data integration offers much stronger generalization, thereby paving a new path for ship maneuvering motion prediction.

Most recently, deep learning-based visual detection has attracted rapidly increasing attention paid to marine organisms, thereby expecting to significantly benefit ocean ecology. Suffering from underwater visual degradation including low contrast, color distortion and blur, etc. , both advances and challenges on visual detection of marine organisms (VDMO) co-exist in the literature. In this survey, deep learning-based VDMO techniques are comprehensively revisited from a systematic viewpoint covering advances in underwater image preprocessing, deep learning-based detection approaches, benchmark dataset and intensively quantitative comparisons. Furthermore, in terms of inherent features of marine organisms and complexity of underwater visual environments, underlying challenges are unfolded in depth. Such a self-contained survey is expected to exploit potential breakthroughs and explore probable trends in deep learning-based VDMO techniques.

The unmanned surface vehicle (USV) has attracted more and more attention because of its basic ability to perform complex maritime tasks autonomously in constrained environments. However, the level of autonomy of one single USV is still limited, especially when deployed in a dynamic environment to perform multiple tasks simultaneously. Thus, a multi-USV cooperative approach can be adopted to obtain the desired success rate in the presence of multi-mission objectives. In this paper, we propose a cooperative navigating approach by enabling multiple USVs to automatically avoid dynamic obstacles and allocate target areas. To be specific, we propose a multi-agent deep reinforcement learning (MADRL) approach, i.e., a multi-agent deep deterministic policy gradient (MADDPG), to maximize the autonomy level by jointly optimizing the trajectory of USVs, as well as obstacle avoidance and coordination, which is a complex optimization problem usually solved separately. In contrast to other works, we combined dynamic navigation and area assignment to design a task management system based on the MADDPG learning framework. Finally, the experiments were carried out on the Gym platform to verify the effectiveness of the proposed method.

In this paper, sufficiently addressing power take-off dynamics, high-accuracy tracking problem of the direct -drive wave energy converter (DWEC) which is inevitably disturbed by complex unknowns is solved by creating a finite-time cascade-like tracking control (FCTC) scheme. More specifically, a cascade structure of the DWEC system is initially built by finely taking d-axis current dynamics into account such that accurate tracking of wave energy can be hierarchically accommodated by independently designing cascade-like control laws for d-and q-voltages. Then, to exactly compensate complex unknowns including unmodeled dynamics and disturbances, a finite-time observer is devised within the foregoing cascade structure, and facilitates the cascade-like controller synthesis whereby the backbone can be rationally decoupled. Within the entire FCTC scheme, d-axis current tracking dynamics can be sufficiently addressed so as to significantly enhance tracking accuracy of wave energy. Eventually, rigorous analysis ensures that both wave displacement and velocity tracking errors are globally exponentially stable. Simulation results and comparisons with typical methods demonstrate remarkable performance and superiority of the proposed FCTC scheme.

Efficient identification of multi-ship encounter situation concerning action priority analysis is of vital significance for making effective and practical collision avoidance manoeuvres. However, action priority analysis is strongly involved in conflict urgency quantification, collision candidates relevance analysis as well as the contribution analysis within the encountering ships. In this paper, considering Maritime Autonomous Surface Ship, a deterministic collision avoidance decision-making system is established to estimate multi-MASS encounter situation. To this end, the approach index and asymmetrical Gaussian fitting method are deployed to assess collision risk, while the encountering ships are analytically distinguished into different clusters based on the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. By virtue of the improved Sharpley value method, the collision avoidance action priority is elaboratively sorted for different clusters. Accordingly, each individual collision avoidance manoeuvres are collaboratively generated by the modified velocity obstacle algorithm with certain time delay. Eventually, the proposed decision-making system is synthesized by functional modules including data-processing, conflict assessment detection, relevance analysis, action priority analysis, path planning and performance monitor. Simulation results demonstrate that this proposed decision-making system can perform significant superiority in various maritime environment in line with the practice of coordination and navigation.

该文通过分析"双一流"背景下高校发展的优势、劣势、机遇与挑战,揭示我国高校师资队伍在教学理念、教学方法、任聘制度等方面存在的问题,从改革教学理念、提升教学水平、建立"高质量代表作"评价体系、建立科学合理的人才任聘考核机制等方面提出建议.

以质量和绩效为导向，完善教师分类评价机制，对于推动促进我国高等院校的教育体制深化和改革，提升教师工作的热情具有重大意义。通过分析当前我国高校教师分类评价机制存在的问题，提出完善针对性建议。

A new evolution method is proposed in this paper to estimate inland river channel transit capacity. Compared with the conventional methods which usually use Fujii ellipsoid ship domain model, the dynamic quaternion ship domain (DQSD) model is used to determine the safe region of the ships sailing in the inland river channel, which takes ship dynamics, human reliability and circumstance factor into account. Fuzzy logics also are applied to obtain the human factors of DQSD model from the questionnaire investigating data. The transit capacities from Yingong island channel data show that the new evaluation method can reduce the conservation of the transit capacity estimation.

Based on ship maneuverability and human reliability models, a novel dynamic quaternion ship domain (DQSD) model is firstly proposed by effectively combining subsystems of ship dynamics, human performance and navigation environment, which results in the quantitative and analytical study on ship domain models. Specifically, the mathematic model group (MMG) ship motion model is used to establish the ship sub-model for estimation of scale parameters in the DQSD model. In order to identify the shape parameters of the DQSD model, the human reliability model is proposed to implement the novel human sub-model for ship navigators, whereby the variables of professional skill level, physical state, mental state and navigation environment disturbance are sufficiently considered as input states. As a consequence, the resultant DQSD model incorporated by ship dynamics and human performance sub-models could be able to realize analytical investigations on ship domains since the factors of ship, human and navigation environment are effectively quantized. Finally, systematic simulation studies and comparative analysis are conducted on three typical ships to demonstrate the effectiveness of the proposed DQSD model.

The vibration control platform is designed involving periodical micro- vibration in Three Dimension (3D) scanning system. The system is identified utilizing neural network based on Levenberg - Marquardt (L-M) algorithm. The L-M algorithm is the combination of the steepest decent algorithm with the Gauss - Newton algorithm so that it has the faster speed of convergence and higher approach accuracy. Compared with the rational BP, the simulation result showed the LM algorithm with learning rate speed up learning process, reduce training time with improve identification accuracy greatly. The identification effect is good.

We propose a Generalized Online Self-constructing Fuzzy Neural Network (GOSFNN) which extends the ellipsoidal basis function (EBF) based fuzzy neural networks (FNNs) by permitting input variables to be modeled by dissymmetrical Gaussian functions (DGFs). Due to the flexibility and dissymmetry of left and right widths of the DGF, the partitioning made by DGFs in the input space is more flexible and more interpretable, and therefore results in a parsimonious FNN with high performance under the online learning algorithm. The geometric growing criteria and the error reduction ratio (ERR) method are incorporated into structure identification which implements an optimal and compact network structure. The GOSFNN starts with no hidden neurons and does not need to partition the input space a priori. In addition, all free parameters in premises and consequents are adjusted online based on the Extended Kalman Filter (EKF) method. The performance of the GOSFNN paradigm is compared with other well-known algorithms like ANFIS, OLS, GDFNN, SOFNN and FAOS-PFNN, etc., on a benchmark problem of multi-dimensional function approximation. Simulation results demonstrate that the proposed GOSFNN approach can facilitate a more powerful and parsimonious FNN with better performance of approximation and generalization.

In this paper, we propose a Generalized Online Self-organizing Fuzzy Neural Network (GOSFNN) for nonlinear dynamic system identification. The GOSFNN extends the ellipsoidal basis function (EBF)-based fuzzy neural networks (FNNs) by permitting input variables to be modeled by dissymmetrical Gaussian functions (DGFs). Due to the flexibility and dissymmetry of left and right widths of the DGF, the partitioning made by DGFs in the input space is more flexible and more economical, and therefore results in a parsimonious FNN with high performance under the online learning algorithm. The geometric growing criteria and the error reduction ratio (ERR) method are used as rule growing strategies to realize the structure learning algorithm which implements an optimal and compact network structure. The proposed GOSFNN starts with no hidden neurons and does not need to partition the input space a priori. In addition, all the free parameters in premises and consequents are online adjusted by using the Extended Kalman Filter (EKF) approach. The performance of the proposed GOSFNN paradigm is compared with other well-known algorithms like OLS, RBF-AFS, DFNN, GDFNN and FAOS-PFNN, etc., on a benchmark problem in the field of nonlinear dynamic system identification. Simulation results demonstrate that the proposed GOSFNN approach would be able to facilitate a more powerful and more economical fuzzy neural network with better identification performance.

In this paper, we propose a novel ship domain model identified by the Fast and Accurate Online Self-organizing Parsimonious Fuzzy Neural Network (FAOS-PFNN), which is an effective and powerful algorithm for nonlinear system identifications. The blocking area is introduced to be the reference model of ship domains to generate testing and checking databases for online modeling based on the FAOS-PFNN. The main features of our proposed method are as follows: (1) a mass of reasonable input-output data pairs possessing the complex nonlinear dynamics of ship domains could be randomly extracted; (2) based on the dependable databases, the intelligent ship domain model could be online identified by the FAOS-PFNN while training data pairs sequentially arrives; (3) dynamic and static parameters of own and target ships encountered could be reasonably and comprehensively incorporated into the resulting fuzzy neural network model of ship domains; and, (4) the shape and size of ship domains could be implemented by three independent fuzzy neural systems based on the FAOS-PFFN. It is shown that the identified ship domain model could capture well the key nonlinear properties of ship domains over a wide range. Simulation studies demonstrate the high performance of identification and generalization in the proposed intelligent ship domain model.

The paper summarized the main research results and corresponding characteristics of types of research methods based on the classification review and analysis of ship domain models, and the main problems were investigated to demonstrate the essence and meaning of study on ship domain, and then several theoretical and practical possible research areas of ship domain are definitely proposed.

后金融危机时期,国际服务外包产业迅速复苏,升级加快和多元发包的特点为中国实现服务外包跨越式发展、服务业水平全面提高和就业结构进一步优化提供了难得的机遇。本文从政府、行业协会和企业三个层面提出改善中国外包软环境、提升外包综合实力的建议。