This paper studies the synchronization problem of two-player multiagent systems through reinforcement learning methods. A Nash-minmax strategy is formulated, where the interactions of two players in the same agent are...
详细信息
ISBN:
(数字)9798350363012
ISBN:
(纸本)9798350363029
This paper studies the synchronization problem of two-player multiagent systems through reinforcement learning methods. A Nash-minmax strategy is formulated, where the interactions of two players in the same agent are non-zero-sum, while interactions of players between agents are zero-sum games. We propose an offline model-based reinforcement learning algorithm to identify Nash solutions for players within each agent, as well as the worst control solutions for players in neighboring antagonistic agents. On this basis, a data-driven off-policy algorithm is provided to alleviate the requirement for accurate system dynamics in the offline algorithm. Besides, the convergence of the proposed algorithms is analyzed. Finally, simulation results verify the effectiveness of the designed algorithms.
Malaria is a critical health condition that affects both sultry and frigid region worldwide, its rise to millions of cases disease. The malaria is caused by parasites through to pass in the human blood cells, damage t...
详细信息
ISBN:
(数字)9798350352931
ISBN:
(纸本)9798350352948
Malaria is a critical health condition that affects both sultry and frigid region worldwide, its rise to millions of cases disease. The malaria is caused by parasites through to pass in the human blood cells, damage them over the time and occur the noise because it blends in the human blood cells. Therefore, it is identifying by using microscope Image to exanimated the blood cells. This paper, proposed Boosting Reweighting Convolutional Neural Network (BR-CNN) technique efficiently analysis the region to malaria parasites classify the disease image. Data obtained from the NIH Malaria dataset and pre-processing utilized remove unwanted noise in the image and normalization of the feature vector. The feature extraction involved the ResNet 50 utilized residual connections, which allows gradients to flow through the network, effectively extracted the malaria parasites. The Classification utilized BR-CNN technique is automatically extract features from the input blood cell image. The CNN processes the image through layer of the convolutional operations and non-linear activation to build a hierarchical representation of data. The performance of proposed model evaluated the various metrics such as accuracy, precision, recall and sensitivity metrics on NIH dataset.
In this paper, we focus on a reconfigurable intelligent surface (RIS)-assisted millimeter-wave (mmWave) communications system, where a RIS is deployed to create reliable reflection links and alleviate multi-user inter...
In this paper, we focus on a reconfigurable intelligent surface (RIS)-assisted millimeter-wave (mmWave) communications system, where a RIS is deployed to create reliable reflection links and alleviate multi-user interference. In the system, each user equipment (UE) is concerned with the freshness of the information received from the base station (BS) and has a certain transmission requirement. The age of information (AoI) is adopted as the metric for information freshness and on this basis, we define the weighted AoI to take into account the satisfaction of UE transmission requirements. Then, an average weighted AoI minimization problem is formulated by jointly optimizing the precoding, the discrete RIS phase shifts, and the scheduling strategy. The problem is decomposed into two subproblems. For the first precoding and RIS design subproblem, we aim to satisfy the transmission requirements of the scheduled UEs with minimum power consumption under a given scheduling strategy. We use the zero-force (ZF)-based precoder and adopt the local search method for the RIS design. For the second scheduling strategy design subproblem, we propose a heuristic scheduling algorithm to minimize the system's average weighted AoI. Simulation results verify the effectiveness of our proposed method over several baseline schemes.
Grid-forming converters are widely envisioned to be the cornerstone of future converter-dominated power systems. However, standard grid forming (GFM) control strategies assume a fully controllable source with enough h...
Grid-forming converters are widely envisioned to be the cornerstone of future converter-dominated power systems. However, standard grid forming (GFM) control strategies assume a fully controllable source with enough headroom behind the converter and only implicitly address renewable generation limits through the converter limits. This can lead to instabilities on time scales of both primary and secondary frequency control and jeopardize the safe and reliable operation of electric power systems. In this work, we leverage the recently proposed dual-port GFM control that maps power imbalances in the grid to the power generation interfaced by the power converter. We show that this mechanism allows for considering and transparently addressing limits of renewable generation (e.g., solar photovoltaics and wind) in primary and secondary frequency control. We illustrate that renewable generation using dual-port GFM control can directly integrate into prevailing secondary control methods such as automatic generation control (AGC). Moreover, we discuss the limitations of standard AGC when one or more areas of a power system are dominated by renewable generation and propose an anti-windup strategy to address the power generation limits of renewables. Finally, we verify our findings in a time-domain, electromagnetic transient (EMT) simulation.
This paper introduces a novel compliant mechanism combining lightweight and energy dissipation for aerial physical interaction. Weighting 400 g at take-off, the mechanism is actuated in the forward body direction, ena...
详细信息
The paper introduces a novel framework for safe and autonomous aerial physical interaction in industrial settings. It comprises two main components: a neural network-based target detection system enhanced with edge co...
详细信息
ISBN:
(数字)9798350357882
ISBN:
(纸本)9798350357899
The paper introduces a novel framework for safe and autonomous aerial physical interaction in industrial settings. It comprises two main components: a neural network-based target detection system enhanced with edge computing for reduced onboard computational load, and a control barrier function (CBF)-based controller for safe and precise maneuvering. The target detection system is trained on a dataset under challenging visual conditions and evaluated for accuracy across various unseen data with changing lighting conditions. Depth features are utilized for target pose estimation, with the entire detection framework offloaded into low-latency edge computing. The CBF-based controller enables the UAV to converge safely to the target for precise contact. Simulated evaluations of both the controller and target detection are presented, alongside an analysis of real-world detection performance.
An optimal modulation scheme with triple-phase-shift (TPS) control could increase the efficiency in the entire load range for a dual-active-bridge (DAB) converter under wide output voltage range conditions. Therefore,...
详细信息
The paper introduces a novel framework for safe and autonomous aerial physical interaction in industrial settings. It comprises two main components: a neural network-based target detection system enhanced with edge co...
详细信息
For doubly salient reluctance machines, the magnetic saturation effect easily occurs, thereby the saturated inductance with complex analytical expression brings difficulties to the inductance modeling procedure during...
详细信息
ISBN:
(数字)9798331528096
ISBN:
(纸本)9798331528102
For doubly salient reluctance machines, the magnetic saturation effect easily occurs, thereby the saturated inductance with complex analytical expression brings difficulties to the inductance modeling procedure during position sensorless control. To tackle this issue, a direct solution from electrical machine design is introduced to regulate magnetic flux distribution. The key is to insert tangentially-magnetized-slot-permanent-magnet into stator slots, thus relieving the magnetic saturation in the stator teeth. Consequently, unsaturated inductance characteristics can be acquired through the magnetic saturation relieving effect, thereby the inductance modeling effort can be significantly decreased, and such design is friendly for doubly salient reluctance machine position sensorless drive application.
This paper presents a learning-based methodology for developing an optimal lane-changing control policy for a Remote controlled (RC) car using real-time sensor data. The RC car is equipped with sensors including GPS, ...
This paper presents a learning-based methodology for developing an optimal lane-changing control policy for a Remote controlled (RC) car using real-time sensor data. The RC car is equipped with sensors including GPS, IMU devices, and a camera integrated in an Nvidia Jetson AGX Xavier board. By a novel Adaptive Dynamic Programming (ADP) algorithm, our RC car is capable of learning the optimal lane-changing strategies based on the real-time processed measurement from the sensors. The experimental outcomes show that our learning-based control algorithm can be effectively implemented, adapt to parameter changes, and complete the lane changing tasks in a short learning time with satisfactory performance.
暂无评论