Optimal Operator Research Articles

De-Pessimism Offline Reinforcement Learning via Value Compensation.

Offline reinforcement learning (RL) has been widely used in practice due to its efficient data utilization, but it still faces the challenge of training vulnerability caused by policy deviation. Existing offline RL methods that add policy constraints or perform conservative Q -value estimation are pessimistic, making the learned policy suboptimal. In this article, we address the pessimism problem by focusing on accurate Q -value estimation. We propose the de-pessimism (DEP) operator to estimate Q values using the optimal Bellman operator or the compensation operator according to whether the actions are in the behavior support set. The compensation operator qualitatively determines the positive or negative nature of out-of-distribution (OOD) actions based on their performance compared with the behavior actions. It leverages differences in state values to compensate for the Q value of positive OOD actions, thereby alleviating pessimism. We theoretically demonstrate the convergence of DEP and its effectiveness in policy improvement. To further advance the practical application, we integrate DEP into the soft actor-critic (SAC) algorithm, yielding the value-compensated de-pessimism offline RL (DoRL-VC). Experimentally, DoRL-VC achieves state-of-the-art (SOTA) performance across mujoco locomotion, Maze 2-D, and challenging Adroit tasks, illustrating the efficacy of DEP in mitigating pessimism.

On the identification and optimization of nonsmooth superposition operators in semilinear elliptic PDEs

We study an infinite-dimensional optimization problem that aims to identify the Nemytskii operator in the nonlinear part of a prototypical semilinear elliptic partial differential equation (PDE) which minimizes the distance between the PDE-solution and a given desired state. In contrast to previous works, we consider this identification problem in a low-regularity regime in which the function inducing the Nemytskii operator is a-priori only known to be an element of H1loc(ℝ). This makes the studied problem class a suitable point of departure for the rigorous analysis of training problems for learning-informed PDEs in which an unknown superposition operator is approximated by means of a neural network with nonsmooth activation functions (ReLU, leaky-ReLU, etc.). We establish that, despite the low regularity of the controls, it is possible to derive a classical stationarity system for local minimizers and to solve the considered problem by means of a gradient projection method. The convergence of the resulting algorithm is proven in the function space setting. It is also shown that the established first-order necessary optimality conditions imply that locally optimal superposition operators share various characteristic properties with commonly used activation functions: They are always sigmoidal, continuously differentiable away from the origin, and typically possess a distinct kink at zero. The paper concludes with numerical experiments which confirm the theoretical findings.

Study on the Equivalent Conversion and Utilisation Model of Solar Energy and its Benefits

Relying on the energy local area network (LAN), we study the equivalent conversion relationship between different “energy qualities” in the process of solar energy conversion and utilisation and its benefit model. Considering the impact of load demand characteristics and current energy consumption price factors on solar energy utilization, a multi-objective optimization model of solar energy conversion and utilization is constructed with the goals of maximizing solar energy utilization and maximizing the comprehensive benefits of system operation; then, the particle swarm optimization algorithm with quantum behaviour is used to optimize and solve the model, and a comparative analysis is carried out with the traditional single-energy supply mode. The results show that the proposed model can effectively improve the comprehensive utilisation and consumption of solar energy by the energy consumption system, and achieve higher energy sales benefits. Therefore, with the help of energy storage technology, the fluctuation and intermittency of the solar PV system output can be suppressed, and flexible grid connection can be realised, so as to enhance the grid-connected capacity of solar energy; at the same time, higher economic benefits can be obtained with the help of optimisation, coordination and control system operation. Optimised control of solar PV power generation systems under different scenarios ensures the safety and stability of the system’s power output and improves the grid-connected capability of solar PV power generation systems. The validity and feasibility of the proposed model and its operation mode are verified, providing a new idea for the development and utilisation of large-scale solar energy.

An Intelligent Edge Dual-Structure Ensemble Method for Data Stream Detection and Releasing

Edge intelligence is a critical enabler of intelligent application services in the Internet of Things (IoT). However, due to complex environmental factors, edge devices are subject to constant dynamic changes, which can result in security threats and sensitive information leakage. Therefore, it is essential to investigate data stream online analysis and detection strategies and implement an online releasing mechanism to ensure sensitive information is not leaked. Existing work rarely addresses these issues simultaneously or has poor performance, which poses a challenge. To address this challenge, we propose an intelligent edge dual-structure ensemble method (IEDSEM), consisting of three key components: data preprocessing, drift detection data analytics (IEDSEM-DDDA), and privacy-preserving data releasing (IEDSEM-PPDR). Data preprocessing is used primarily to enhance the quality of data streams to improve the performance of model learning. IEDSEM-DDDA involves three sequential operations: dynamic feature selection, model learning and selection, and online model ensemble deployment to achieve anomaly detection of online data streams. Meanwhile, IEDSEM-PPDR uses differential privacy and online optimization operations to achieve intelligent hierarchical protection of edge data. To validate the performance of our proposed IEDSEM method, we conducted two comprehensive simulation experiments on real data machines, verifying the accuracy of the concept drift component detection and the privacy optimization performance of the privacy-preserving component, respectively. Simulation results show that compared with several other advanced high-performance algorithms, our algorithm has over 99% accuracy in data stream analysis detection and more outstanding privacy-preserving ability.

OPTIMALISASI PENGELOLAAN EKOWISATA HUTAN MANGROVE “LATI TUO“ UNTUK MENINGKATKAN PEREKONOMIAN WARGA DESA KLEMPANG SARI

Tourism, including environmental tourism, has become one of Indonesia's second-greatest foreign exchange contributors after petroleum and natural gas. The potential for Mangrove tourism is growing due to the ecotourism philosophy, which stresses nature conservation and local communities, particularly coastal areas. The solution to the issue of developing tourism activities in conservation areas is ecotourism. There is a Mangrove Forest ecotourism area called "Lati Tuo" in Klempang Sari Village, Kuaro District, Paser Regency, however, its management is not yet optimal. This activity intends to improve the management of "Lati Tuo" Mangrove ecotourism by developing supporting infrastructure at ecotourism sites. All activities involved teams from the village, Mulawarman University students, ecotourism managers, and the community around the location. The actions carried out include preparation, problem identification, planning, implementation, and assessment. Activities carried out at the "Lati Tuo" mangrove ecotourism site, Klempang Sari Village. This time, ecotourism management optimization operations are more focused on building several supporting facilities for visitor comfort, because tourist comfort is one of the variables that might influence tourists' decisions to visit a tourist site. Making chairs at each corner of the bridge and creating warning signs are among the activities that have taken place. Aside from that, there was the creation of gazebo name boards, the creation of Latin name plates for plants in the mangrove forest "Lati Tuo," and the creation of "Lati Tuo" Love Padlocks. The development of supporting facilities at the "Lati Tuo" mangrove ecotourism site was completed smoothly. The support of all stakeholders, including the village officials, ecotourism management, and the Klempang Sari Village Community, is critical to the success of this activity.

Wave function of the sexaquark or compact H-dibaryon

We derive and explicitly display the internal wave function of a color-, flavor- and spin- singlet dibaryon composed of uuddss quarks in a spatially symmetric state, in the approximation of exact SU(3) flavor symmetry. This wavefunction shows that the often-used superposition of ΛΛ, NΞ and ΣΣ baryons, relevant for di-baryon molecules, accounts for only 1/5 of a spatially symmetric six-quark color-flavor-spin-singlet state, with the remaining 4/5 consisting of products of color-octet baryons. Using the correct wavefunction has important implications for calculations of the mass of the lowest-lying flavor-singlet dibaryon, as we illustrate using the Cornell potential. We also provide a compact representation of the state in terms of creation operators, and comment on the impact of not using the optimal operator in lattice studies of this system.

Multi-class classification of control room operators’ cognitive workload using the fusion of eye-tracking and electroencephalography

Chemical process industries are hazard intensive. Most industrial accidents occur today due to human error. For safe and efficient operation, it is therefore critical to ensure optimal operator performance. With the advent of Industry 4.0 and concomitant digitalization, the role of operators has become cognitively challenging. Therefore, it is imperative to assess the cognitive performance of operators. One of the major constructs to understand cognitive performance is the cognitive workload. An increase in cognitive workload often leads to degradation in performance. Traditional assessment techniques fail to capture cognitive aspects of performance. Recently, researchers in various domains such as aviation, driving, marine, and nuclear power have started to utilize physiological measures to gauge the cognitive workload of their operators. In our previous works, we have used electroencephalography (EEG) and eye-tracking separately to assess the cognitive workload of operators in the process industries. In contrast, in this paper, we explore the benefits of their fusion in classifying process industry control room operators’ workload into low, medium, and high classes while they tackle abnormal situations. The methodology employs the fusion of metrics derived from pupil, gaze, and EEG data to train a decision tree-based model for workload classification. Our results reveal that fusion leads to an increase in classification accuracy of upto 22 %. The work has the potential to identify the expertise level of operators and hence, can be critical in ensuring their optimal performance.

The Performance of Reinforcement Learning for Indoor Climate Control Devices according to the Level of Outdoor Air Particulate Matters

As people spend more than 90% of their time indoors, indoor environmental quality (IEQ) has become an important factor in maintaining a healthy space for the occupants. There are many indoor climate control devices for improving IEQ. However, it is difficult to maintain an appropriate IEQ with changing outdoor air conditions and occupant behavior in a building. This study proposes a reinforcement learning (RL) algorithm to maintain indoor air quality (IAQ) with low energy consumption in a residential environment by optimally operating indoor climate control devices such as ventilation systems, air purifiers, and kitchen hoods. The proposed artificial intelligence algorithm (AI2C2) employs DDQN (double deep Q-network) to determine the optimal operation of various indoor climate control devices, reflecting IAQ and energy consumption via different outdoor levels of particulate matter. This approach considers the outdoor air condition and occupant activities in training the developed algorithm, which are the most significant factors affecting IEQ and building energy performance. A co-simulation platform using Python and EnergyPlus is applied to train and evaluate the model. As a result, the proposed approach reduced energy consumption and maintained good IAQ. The developed RL algorithm for energy and IAQ showed different performances based on the outdoor PM 2.5 level. The results showed the RL-based control can be more effective when the outdoor PM 2.5 level is higher (or unhealthy) compared to moderate (or healthy) conditions.

Quantum correlation functions through tensor network path integral.

Tensor networks have historically proven to be of great utility in providing compressed representations of wave functions that can be used for the calculation of eigenstates. Recently, it has been shown that a variety of these networks can be leveraged to make real time non-equilibrium simulations of dynamics involving the Feynman-Vernon influence functional more efficient. In this work, a tensor network is developed for non-perturbatively calculating the equilibrium correlation function for open quantum systems using the path integral methodology. These correlation functions are of fundamental importance in calculations of rates of reactions, simulations of response functions and susceptibilities, spectra of systems, etc. The influence of the solvent on the quantum system is incorporated through an influence functional, whose unconventional structure motivates the design of a new optimal matrix product-like operator that can be applied to the so-called path amplitude matrix product state. This complex-time tensor network path integral approach provides an exceptionally efficient representation of the path integral, enabling simulations for larger systems strongly interacting with baths and at lower temperatures out to longer time. The derivation, design, and implementation of this method are discussed along with a wide range of illustrations ranging from rate theory and symmetrized spin correlation functions to simulation of response of the Fenna-Matthews-Olson complex to light.

Research of a Multi-Level Organization Human Resource Network Optimization Model and an Improved Late Acceptance Hill Climbing Algorithm

Complex hierarchical structures and diverse personnel mobility pose challenges for many multi-level organizations. The difficulty of reasonable human resource planning in multi-level organizations is mainly caused by ignoring the hierarchical structure. To address the above problems, firstly, a multi-level organization human resource network optimization model is constructed by representing the turnover situation of multi-level organizations in a dimensional manner as a multi-level network. Secondly, we propose an improved late acceptance hill climbing based on tabu and retrieval strategy (TR-LAHC) and designed two intelligent optimization operators. Finally, the TR-LAHC algorithm is compared with other classical algorithms to prove that the algorithm provides the best solution and can effectively solve the personnel mobility planning problem in multi-level organizations.

A deep reinforcement learning based algorithm for a distributed precast concrete production scheduling

The environmental-friendly production demands higher manufacturing efficiency and lower energy cost; therefore, time-of-use electricity price constraint and distributed production have attracted more attention. For concrete precast architecture construction, the tardiness penalty and warehouse cost cannot be ignored, which should be optimized for lower cost. In this study, the concrete precast process is investigated as the group scheduling of a distributed flexible job shop problem. Each concrete precast is managed in a group, the setup time between different groups is considered. Two objectives, total weighted earliness and tardiness and total time-of-use electricity cost are minimized, simultaneously. To solve the integrated problem, three coordinated double deep Q-networks (DQN) are applied, which are organized as a learn-to-improve reinforcement learning approach. For distributed scheduling problem, operators in a single factory or in multiple factories differs in solution improvement; so selection DQN is designed to decide the type of the operators according to scheduling circumstances. Other two DQNs, i.e., local DQN and global DQN, are to select the optimization operators in one factory or in multiple factories, respectively. Furthermore, two solution refinement strategies are designed to decrease the objectives after reinforcement learning component. Numerical experiment and statistical analysis suggest that the proposed deep reinforcement learning based algorithm has superiority in solving the considered problem.

Implementation of a Reduced Order Model for rotating seal annular leakage flow inside a centrifugal pump

Sustainability and net zero emissions targets drive the importance of turbomachinery design, optimisation, and efficient operation in modern geoenergy systems. To further enable integration of advanced large-scale technologies, accurate and resource-efficient simulations of complex equipment and systems need to be completed readily. An approach to simulate internal leakage flows of centrifugal pumps is presented, an important factor that influences equipment operation efficiency.One significant challenge when internal leakage flow is investigated is achieving sufficient simulation detail in the flow regions separating rotating and stationary components where close-running clearances are present. Computational fluid dynamics (CFD) meshes typically require excessively fine resolutions of node points in these situations for accuracy. For example, the typical annular suction wear ring clearance ranges between 0.1–0.5 mm, whilst impeller diameters and other geometric properties have length scales from 600–1200 mm or even larger.A reduced order model (ROM) of an annular seal configuration was developed that allows for analytical calculation of the leakage flow by implementing user scripting in the CFD solver setup. The ROM implementation has a low symbol count and does not require any type of logic and/or recursive and iterative functionality. Thus, fine and resource-intensive meshing can be robustly eliminated and replaced by a sufficiently detailed ROM that contributes leakage flow effects to the overall simulation.Leakage flow through the suction wear ring of a high differential head (ΔH), high flow rate, double suction centrifugal pump with water as working fluid was used as a case study for ROM development. The calibrated model was verified against a commercial fluid dynamics software package and achieved a level of precision of 2.4% with σ=0.0118 in a wide operating range of 10m⩽ΔH⩽350m and fluid temperature (T) of 283.15K⩽T⩽353.15K.The leakage flow calculated by the ROM was proven to be accurate and therefore capable of accounting for a range of pressure heads, incorporating surface roughness effects and temperature variations, and adaptable for various fluids. Guidance for the implementation of the ROM in the Ansys CFX solver environment is also provided to aid future work.

Optimized Trajectory Tracking for Robot Manipulators with Uncertain Dynamics: A Composite Position Predictive Control Approach

This study addresses the trajectory tracking control challenges of robot manipulators with uncertain dynamics. The aim is to achieve precise and smooth trajectory regulation through a novel composite position predictive control (PPC) scheme that integrates motion profile and disturbance preview techniques. First, we perform offline dynamics identification and feedforward compensation alongside a pre-defined motion profile. To handle the disturbances arising from uncertain dynamics, a super-twisting disturbance observer is designed, resulting in a dynamically compensated prediction model. Furthermore, the receding optimization operations for PPC are executed by solving an optimal solution associated with a joint angle tracking error. The combination of feedforward and feedback control improves the robot manipulator’s absolute positioning accuracy as opposed to the conventional model predictive control method, especially when dealing with uncertain dynamics. The effectiveness of the proposed control method is confirmed through trajectory tracking experiments conducted on a six-degree-of-freedom robot platform with varying end-effector loads. The experimental results demonstrate that the proposed PPC method enhances tracking accuracy by approximately 45% and 25% when compared to the traditional inverse dynamic control (IDC) and the robust IDC approaches, respectively.

Distributed economic optimisation of multi‐energy park operation based on cloud platform architecture considering network delivery capacity

AbstractTo address the issues of node interaction power overrun and high carbon emissions that may arise during distributed optimization in multi‐energy parks (MEPs), this paper proposes a distributed low‐carbon and economic operation method for multi‐energy parks based on a cloud platform that considers network transmission capacity.The proposed method achieves maximun profit by designing a two layer collaborative architecture for distributed optimization operations. At the top level, cloud platform services are utilized to build a model for checking network transport capacity and carbon emission quotas, optimizing network node over‐limit inspection . The bottom layer constructs a distributed optimization model for multi‐energy complementation in each park, taking into account the information privacy and individual interests of each multi‐energy park. An improved alternating direction multiplier method (ADMM) is proposed to effectively solve the two‐layer framework. The case studies show that the distributed optimization method under cloud platform services proposed in this paper can achieve maximum revenue for integrated energy service provider while ensuring the safe operation of multi‐energy parks, and reasonably allocate the benefits of collaborative operation among various parks while promoting carbon emission reduction.

Brunovsky Decomposition for Dynamic Interval Localization

This paper proposes a new set-membership method for estimating the trajectories of dynamical systems, when the states are completely unknown and only non-linear observations are available. The first part of the proposed method is symbolic and follows the decomposition of Brunovsky, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.</i> , it decomposes the set of differential equations describing the dynamical system into two blocks of constraints: one block gathers non-linear analytical equations that do not involve differential operators, while the other block is composed of linear chains of integrators. The second part of the method, that relies on the symbolic decomposition, is numerical and based on a contractor approach. It involves a specific optimal operator for narrowing the sets of feasible solutions. This approach is shown to be efficient on a difficult problem of dynamic localization of a mobile robot, without any prior knowledge about its states.

Robust Peak Detection for Holter ECGs by Self-Organized Operational Neural Networks.

Although numerous R-peak detectors have been proposed in the literature, their robustness and performance levels may significantly deteriorate in low-quality and noisy signals acquired from mobile electrocardiogram (ECG) sensors, such as Holter monitors. Recently, this issue has been addressed by deep 1-D convolutional neural networks (CNNs) that have achieved state-of-the-art performance levels in Holter monitors; however, they pose a high complexity level that requires special parallelized hardware setup for real-time processing. On the other hand, their performance deteriorates when a compact network configuration is used instead. This is an expected outcome as recent studies have demonstrated that the learning performance of CNNs is limited due to their strictly homogenous configuration with the sole linear neuron model. This has been addressed by operational neural networks (ONNs) with their heterogenous network configuration encapsulating neurons with various nonlinear operators. In this study, to further boost the peak detection performance along with an elegant computational efficiency, we propose 1-D Self-Organized ONNs (Self-ONNs) with generative neurons. The most crucial advantage of 1-D Self-ONNs over the ONNs is their self-organization capability that voids the need to search for the best operator set per neuron since each generative neuron has the ability to create the optimal operator during training. The experimental results over the China Physiological Signal Challenge-2020 (CPSC) dataset with more than one million ECG beats show that the proposed 1-D Self-ONNs can significantly surpass the state-of-the-art deep CNN with less computational complexity. Results demonstrate that the proposed solution achieves a 99.10% F1-score, 99.79% sensitivity, and 98.42% positive predictivity in the CPSC dataset, which is the best R-peak detection performance ever achieved.

Multi‐Party Quantum Key Distribution Using Variational Quantum Eigensolvers

AbstractThe research paper introduces a novel approach to multi‐party quantum key distribution using variational quantum eigensolvers (VQEs). The protocol aims to establish secure communication among multiple parties in a quantum network. The paper outlines a comprehensive framework incorporating quantum state preparation, VQE‐based key generation, error correction, and privacy amplification circuits. Mathematical formulations guide the design of quantum gates, variational parameters, and error correction codes. The protocol utilizes quantum entanglement, VQE optimization, and classical operations for error correction and secure key extraction. The paper's experimental focus involves implementing the proposed protocol on a quantum computer, analyzing security metrics such as mutual information and min‐entropy, and simulating circuit outcomes.

Rapid prediction, optimization and design of solar membrane reactor by data-driven surrogate model

Membrane reactor is a process intensification technology that enhances traditional reactions. This study investigates the optimal operating conditions and structural designs of an insert-enhanced solar methane-to-hydrogen Pd-based membrane reactor (SMMR) for more efficient conversion-reaction and separation-purification. Applying response surface methodology, a data-driven surrogate model is fitted to SMMR for rapid performance prediction, which can combine algorithms to optimize operating conditions and structures. The results indicate that operating conditions center around matching reaction kinetics to strengthen reaction, while structural parameters focus on weakening concentration polarization to enhance separation. Methane feed flow and steam-to-methane ratio are almost always the dominant factors, especially in balancing methane conversion and fuel efficiency. Moreover, SMMR under high-pressure more relies on optimized inserts’ transport enhancement to improve overall reaction-separation performance. The optimized results by single- and multi-objective optimization are similar, where the 4-helical insert delivers a comprehensive excellent methane conversion of 0.95, fuel efficiency of 0.86 and hydrogen recovery of 0.95 at 400 °C inlet temperature and 8 bar pressure. Overall, the optimization of SMMR surrogate model consumes less than 0.2 % of the single simulation CPU time and the average error of the results is about 3 %, showing high efficiency and prediction accuracy.

Multi-UAV Data Collection and Path Planning Method for Large-Scale Terminal Access.

In the context of the relentless evolution of network and communication technologies, the need for enhanced communication content and quality continues to escalate. Addressing the demands of data collection from the abundance of terminals within Internet of Things (IoT) scenarios, this paper presents an advanced approach to multi-Unmanned Aerial Vehicle (UAV) data collection and path planning tailored for extensive terminal accessibility. This paper focuses on optimizing the complex interplay between task completion time and task volume equilibrium. To this end, a novel strategy is devised that integrates sensor area partitioning and flight trajectory planning for multiple UAVs, forming an optimization framework geared towards minimizing task completion duration. The core idea of this work involves designing an innovative k-means algorithm capable of balancing data quantities within each cluster, thereby achieving balanced sensor node partitioning based on data volume. Then, the UAV flight trajectory paths are discretely modeled, and a grouped, improved genetic algorithm is used to solve the Multiple Traveling Salesman Problem (MTSP). The algorithm introduces a 2-opt optimization operator to improve the computational efficiency of the genetic algorithm. Empirical validation through comprehensive simulations clearly underscores the efficacy of the proposed approach. In particular, the method demonstrates a remarkable capacity to rectify the historical issue of diverse task volumes among multiple UAVs, all the while significantly reducing task completion times. Moreover, its convergence rate substantially outperforms that of the conventional genetic algorithm, attesting to its computational efficiency. This paper contributes an innovative and efficient paradigm to improve the problem of data collection from IoT terminals through the use of multiple UAVs. As a result, it not only augments the efficiency and balance of task distribution but also showcases the potential of tailored algorithm solutions for realizing optimal outcomes in complex engineering scenarios.

A Ground-Risk-Map-Based Path-Planning Algorithm for UAVs in an Urban Environment with Beetle Swarm Optimization

This paper presents a path-planning strategy for unmanned aerial vehicles (UAVs) in urban environments with a ground risk map. The aim is to generate a UAV path that minimizes the ground risk as well as the flying cost, enforcing safety and efficiency over inhabited areas. A quantitative model is proposed to evaluate the ground risk, which is then used as a risk constraint for UAV path optimization. Subsequently, beetle swarm optimization (BSO) is proposed based on a beetle antennae search (BAS) that considers turning angles and path length. In this proposed BSO, an adaptive step size for every beetle and a random proportionality coefficient mechanism are designed to improve the deficiencies of the local optimum and slow convergence. Furthermore, a global optimum attraction operator is established to share the social information in a swarm to lead to the global best position in the search space. Experiments were performed and compared with particle swarm optimization (PSO), genetic algorithm (GA), firefly algorithm (FA), and BAS. This case study shows that the proposed BSO works well with different swarm sizes, beetle dimensions, and iterations. It outperforms the aforementioned methods not only in terms of efficiency but also in terms of accuracy. The simulation results confirm the suitability of the proposed BSO approach.