Real-world optimization problems often involve multiple (typically conflicting) objective functions that must be balanced since they cannot reach their optima simultaneously. Interactive multiobjective optimization me...
详细信息
ISBN:
(数字)9798331520151
ISBN:
(纸本)9798331520168
Real-world optimization problems often involve multiple (typically conflicting) objective functions that must be balanced since they cannot reach their optima simultaneously. Interactive multiobjective optimization methods are powerful in solving such problems because they incorporate the preference information of a domain expert, a decision maker (DM), in the solution process to find the most preferred trade-offs. However, the connection between the preferences and the solutions generated based on them may remain unclear, limiting a DM’s understanding and confidence in the solution process. Therefore, we propose a novel approach leveraging local interpretable model-agnostic explanations (LIME) to make interactive methods *** generate local linear approximations of the set of solutions that reflect different trade-offs to provide a DM with insights into the behavior of objective functions near the solutions considered. We can explain how changes in one objective function value affect others, assisting a DM in adjusting their preferences during the interactive solution process to get more preferred solutions. We utilize Karush-Kuhn-Tucker multipliers to construct the linear model analogous to LIME. This is convenient since the multipliers can readily be obtained as a byproduct of applying an appropriate numerical optimization *** our knowledge, this is the first application of LIME-inspired ideas to the domain of interactive multiobjective optimization. The proposed approach can significantly enhance the transparency and trustworthiness of interactive methods, ultimately leading to better decisions.
Recently, an absolute value inequalities discriminant analysis criterion with robustness and sparseness for supervised dimensionality reduction was studied. However, it obtains discriminant directions one by one throu...
详细信息
Recently, an absolute value inequalities discriminant analysis criterion with robustness and sparseness for supervised dimensionality reduction was studied. However, it obtains discriminant directions one by one through greedy search, which makes the sparseness of multiple discriminant directions unexplainable. In addition, it relaxes the original problem into a series of linear programming problems which makes it time consuming. In this paper, we construct a novel linear discriminant analysis with robustness and sparseness jointly through the $L_linear programming$ -norm and $L_{2,1}$ -norm. The proposed approach obtains all the discriminant directions simultaneously, and rather than solving linear programming problems, it is solved by a more effective alternating direction method of multipliers. The effectiveness of the proposed method is supported by preliminary experimental results on two artificial datasets, some benchmark datasests and two face image datasets.
High permeability of renewable energy into the power system can cause risks such as overvoltage, frequency variation, voltage deviation, reverse power flow (RPF), etc. This paper presents a method to detect and minimi...
详细信息
ISBN:
(数字)9798331506797
ISBN:
(纸本)9798331506803
High permeability of renewable energy into the power system can cause risks such as overvoltage, frequency variation, voltage deviation, reverse power flow (RPF), etc. This paper presents a method to detect and minimize the reverse power flow in a distribution system integrated with renewable energy. IEEE 13 bus with a monitor is mounted on the source bus to identify the reverse power flow used to test the system. To demonstrate the induction of reverse power flow, a PV system is connected at node 611 with a set capacity of 4000 kW. The load flow is performed using co-simulation between engines OpenDSS and MATLAB through component object model server dynamiclink library. Biogeography-based optimization (BBO) is applied to solve the multiple objective functions, including RPF at the substation and total load increase at each bus. Combining BBO and PSO can balance both global and local search capabilities. Results obtained by the proposed method show that the reverse power flow at the substation occurring due to the connection of large amounts of renewable energy sources can be significantly reduced.
This work studies joint waveform design and subcarrier allocation in a multiuser multicarrier monostatic joint communications and sensing (JCAS) system. We aim to design the JCAS transmit waveform with subcarrier allo...
详细信息
ISBN:
(数字)9798350368369
ISBN:
(纸本)9798350368376
This work studies joint waveform design and subcarrier allocation in a multiuser multicarrier monostatic joint communications and sensing (JCAS) system. We aim to design the JCAS transmit waveform with subcarrier allocation that maximizes the energy efficiency (EE) of the multiuser communications subject to the constraints on the minimum communications and sensing signal-to-interference-plus-noise ratios (SINRs) and transmit power budget. The formulated EE design problem is a mixed-integer non-convex fractional program, which is highly challenging to solve directly. To overcome the challenges, we leverage alternating optimization (AO) and divide the original problem into two subproblems, namely, the waveform design and subcarrier allocation. For the non-convex fractional objective function in the waveform design, we propose an efficient method combining two typical techniques in fractional programming, namely, the quadratic and Dinkelbach transforms, along with successive convex approximation (SCA). For the subcarrier allocation, we leverage quadratic transform and penalty function method. Numerical results demonstrate the effectiveness of the proposed method, showing a significant improvement in energy efficiency compared to a baseline scheme.
Feature descriptors play a crucial role in image retrieval by representing images in a compact and discriminative manner. This survey paper explores various techniques for preparing feature descriptors used in image r...
详细信息
ISBN:
(数字)9798331523893
ISBN:
(纸本)9798331523909
Feature descriptors play a crucial role in image retrieval by representing images in a compact and discriminative manner. This survey paper explores various techniques for preparing feature descriptors used in image retrieval systems. We categorize the approaches into traditional handcrafted descriptors and deep learning-based descriptors. For each category, we discuss popular techniques, their underlying principles, and their advantages and limitations. We also provide insights into recent advancements and benchmark datasets used for evaluation. In order to help academics and practitioners choose the best strategies for their applications, this survey attempts to give them a thorough grasp of feature descriptor preparation methods in image retrieval.
Aiming at the problems of uncertainty of wind and photovoltaic power output and disorderly charging of a large number of electric vehicles accessing the distribution network leading to the superposition of distributio...
详细信息
ISBN:
(数字)9798331506797
ISBN:
(纸本)9798331506803
Aiming at the problems of uncertainty of wind and photovoltaic power output and disorderly charging of a large number of electric vehicles accessing the distribution network leading to the superposition of distribution network load peaks and the increase of network loss, a multi-timescale loss-reduction operation strategy of the distribution network considering electric vehicles is proposed. The probability distribution of EV behavior is established, and the EV charging output model conforming to the user's travel habit is built using the Latin hypercubic sampling method. Considering the influence of the deviation of the wind and photovoltaic output forecasts on different time scales, a multi-timescale optimal dispatch strategy based on the day-ahead-intraday is established, and a mathematical model is constructed by introducing the dynamic time-of-day tariff and using the network loss and the operating cost as the objective function. Different scenarios are simulated in the improved IEEE 33-bus system, and the comparative results show that the proposed optimization strategy is effective in reducing losses and smoothing load fluctuations, and the flexibility of EVs participating in the demand response of the grid reduces the impact of the wind and photovoltaic power integration while compensating for the deviation of the forecast of the wind and photovoltaic power outputs, thus maintaining the system balance.
The constrained discrete optimization problems (CDOP) have a stochastic objective function and deterministic inequality constraints. The CDOP is NP-hard due to the large and exponentially growing solution space. The o...
详细信息
ISBN:
(数字)9798331522193
ISBN:
(纸本)9798331522209
The constrained discrete optimization problems (CDOP) have a stochastic objective function and deterministic inequality constraints. The CDOP is NP-hard due to the large and exponentially growing solution space. The ordinal optimization framework is treated as a recognized framework for resolving NP-hard problems. A simheuristic approach cooperating skill optimization algorithm with ordinal optimization is developed to solve this CDOP in a short time. The simheuristic approach comprises two levels. The skill optimization algorithm is used to construct a differentiated subset with a cursory estimate at the first level. Then, the ranking and selection method is used to find a superior solution from differentiated subsets in the second level. Finally, the approach is applied to the multiple-item inventory system's stock level optimization problem. A real-life example with 5 potential customers and 3 products is used to test the performance. Simulation results reveal that the simheuristic approach can obtain stable and excellent results on the CDOP.
Deep reinforcement learning is the combination of reinforcement learning and deep learning. Reinforcement learning is used to define the problem itself and optimize the objective function, and deep learning is used to...
详细信息
ISBN:
(数字)9798331531935
ISBN:
(纸本)9798331531942
Deep reinforcement learning is the combination of reinforcement learning and deep learning. Reinforcement learning is used to define the problem itself and optimize the objective function, and deep learning is used to solve the problems such as state representation and policy expression. The advantage of attention mechanism is the ability to focus on the important parts of the input information. Object detection is an important branch of computer vision. The detection accuracy is constantly improving, but it is still far from meeting the requirements of human beings. One of the main reasons is that the model lacks the ability to express the target to be detected. In order to solve the problem of low detection accuracy and slow speed in deep reinforcement learning object detection model, this paper proposes a deep reinforcement learning object detection model based on improved convolutional attention mechanism. Through experiments, it is found that the model has better performance in the detection task.
We study the design of orthogonal frequency division multiplexing (OFDM) waveform for the joint sensing and communications (JSC), whose sub-carriers are to be optimally coded by a set of sequences. To obtain such wave...
详细信息
ISBN:
(数字)9798350368741
ISBN:
(纸本)9798350368758
We study the design of orthogonal frequency division multiplexing (OFDM) waveform for the joint sensing and communications (JSC), whose sub-carriers are to be optimally coded by a set of sequences. To obtain such waveform that simultaneously exploits frequency diversity and modulation flexibility for the JSC, we take the correlation level of waveform for sensing and the orthogonality of modulation sequences for communications into consideration. Specifically, we choose to minimize the integrated sidelobe level (ISL) of the OFDM waveform, and meanwhile, to ensure reasonable constraints enforced on sub-carriers with code modulations. In view of this, an ISL-minimization based design with respect to the modulation sequences is therefore formulated, which is generally non-convex. To tackle it, we introduce virtually auxiliary variables to help reformulate the original optimization problem, and then apply the framework of consensus alternating direction method of multipliers for finding solutions. Our major contributions also lie in elaborating an augmented Lagrangian for the newly obtained optimization problem via a first-order Taylor expansion on its objective function, based on which a closed-form solution is achieved for iterations. Simulation results verify the superiority of our proposed OFDM waveform design.
In this paper, we consider the computational protein design (CPD) problem, which is usually modeled as 0/1 programming and is extremely challenging due to its combinatorial properties. As a quadratic semi-assignment p...
详细信息
暂无评论