Class Of Optimization Algorithms Research Articles

A Prospect-Theory-Based Operation Loop Decision-Making Method for Kill Web

In the military field, decision making has become the core of the new operational concept, known as the “kill web”. Although the theory of kill web has been widely recognized by many countries, the decision-making methods for the kill web are still in the early stage. Therefore, there is a need for a new decision-making method for the kill web. Firstly, different from the traditional scheme decision, the kill web is a complex system. The method of complex network provides a new perspective on complex systems, so the kill web was modeled based on complex network. Secondly, the kill web relies on artificial intelligence to provide decision-makers with operation loop solutions, and then decision-makers rely on the experience to make a final decision. However, the current decision-making methods only consider one of the intelligent and human decision-making methods, while the kill web needs to consider both. Hence, we combined intelligent decision making with human decision making through multi-objective optimization and the prospect theory. Finally, we designed a nondominated sorting ant colony genetic algorithm-II (NSACGA-II) to solve large-scale problems, since the kill web is a large-scale system. In addition, an illustrative case was used to verify the feasibility and effectiveness of the proposed model. The results showed that, compared with other classical multi-objective optimization algorithms, the NSACGA-II is superior to other superior algorithms in terms of the hypervolume (HV) and spacing (SP), which verifies the effectiveness of the method and greatly improves the quality of commanders’ decision-making.

Optimization of Vehicle Paths considering Carbon Emissions in a Time-Varying Road Network

Regarding the study of the time-dependent green vehicle path problem (TDGVRP), most of the literature uses the step function to represent the vehicle speed change in order to reduce the computation, ignoring the continuity of vehicle speed, which leads to the lack of accurate carbon emissions measurement. This study represents the vehicle speed variation as a continuous function to make the constructed model more consistent with the actual situation, in order to promote the reduction of carbon emissions generated in the logistics and distribution process, improve the greenhouse effect and ecological environment, and ultimately promote sustainable development. In this paper, a simulated annealing-genetic hybrid algorithm (GA-SA) is proposed to solve the constructed optimization model, and two sets of comparison experiments are designed. The experimental results show that compared with the two classical algorithms, the simulated annealing-genetic hybrid algorithm (GA-SA) has better solution performance, inherits the robustness and potential parallelism of the genetic algorithm, and has a higher practical value. Meanwhile, although the total driving distance of the vehicle path considering carbon emissions increases by 3.52 km, the carbon emission cost and the total cost decrease by 5.6% and 3.4%, respectively, which confirms that the path optimization model considering carbon emissions constructed in this study can not only play the role of restraining carbon emissions but also reduce the total distribution cost and the waste of resources. In this study, a continuous function is used to represent the vehicle speed variation, and two classical optimization algorithms (the genetic algorithm and simulated annealing algorithm) are combined and parameter-optimized, and certain innovations are made in the processing of vehicle speed and the solution algorithm. Finally, the effectiveness of the model and algorithm is verified by experiments.

A Personalized User Authentication System Based on EEG Signals.

Conventional biometrics have been employed in high-security user-authentication systems for over 20 years now. However, some of these modalities face low-security issues in common practice. Brainwave-based user authentication has emerged as a promising alternative method, as it overcomes some of these drawbacks and allows for continuous user authentication. In the present study, we address the problem of individual user variability, by proposing a data-driven Electroencephalography (EEG)-based authentication method. We introduce machine learning techniques, in order to reveal the optimal classification algorithm that best fits the data of each individual user, in a fast and efficient manner. A set of 15 power spectral features (delta, theta, lower alpha, higher alpha, and alpha) is extracted from three EEG channels. The results show that our approach can reliably grant or deny access to the user (mean accuracy of 95.6%), while at the same time poses a viable option for real-time applications, as the total time of the training procedure was kept under one minute.

An Effective Fault Diagnosis Technique for Wind Energy Conversion Systems Based on an Improved Particle Swarm Optimization

The current paper proposes intelligent Fault Detection and Diagnosis (FDD) approaches, aimed to ensure the high-performance operation of Wind energy conversion (WEC) systems. First, an efficient feature selection algorithm based on particle swarm optimization (PSO) is proposed. The main idea behind the use of the PSO algorithm is to remove irrelevant features and extract only the most significant ones from raw data in order to improve the classification task using a neural networks classifier. Then, to overcome the problem of premature convergence and local sub-optimal areas when using the classical PSO optimization algorithm, an improved extension of the PSO algorithm is proposed. The basic idea behind this proposal is to use the Euclidean distance as a dissimilarity metric between observations in which a single observation is kept in case of redundancies. In addition, the proposed reduced PSO-NN (RPSO-NN) technique not only enhances the results in terms of accuracy but also provides a significant reduction in computation time and storage cost by reducing the size of the training dataset and removing irrelevant and redundant samples. The experimental results showed the robustness and high performance of the proposed diagnosis paradigms.

Concrete chloride diffusion modelling using marine creatures-based metaheuristic artificial intelligence

Chloride-induced steel reinforcement corrosion endangers the durability of concrete structures and may cause substantial economic loss and environmental impact. An accurate estimation of chloride diffusion of steel-reinforced concrete assists in the reliable prediction of its service life. Artificial neural network (ANN) trained by classical optimization algorithms suffer from the overfitting issue. However, metaheuristic algorithms keep the balance between exploration and exploitation to overcome this issue. Four marine creatures-based metaheuristic optimization algorithms, i.e., Jellyfish search optimizer (JSO), marine predators algorithm (MPA), salp swarm algorithm (SSA), and whale optimization algorithm (WOA) are synthesized by ANN in this study. The proposed algorithms are served to model the apparent chloride diffusion (Dap) of concrete exposed to atmosphere, tidal, splash, and submerged environments. A total of 216 data from relevant field experiments was extracted from the literature. The results indicate that the synthesized algorithms with simpler architectures perform better than the traditional algorithm. The Wilcoxon rank-sum test shows that the ANN-JSO algorithm performs better than other ANN algorithms, and there is no preference among the average performances of the ANN-MPA, ANN-SSA, and ANN-WOA in the testing phase. The global sensitivity analysis denotes that the exposure time, curing conditions, and exposure are the most critical parameters in predicting the Dap of concrete.

High-Generalizability Reinforcement Learning Based Capacity Optimization in WDM Long-Haul Networks

We propose a fast machine learning assisted capacity optimization in a point-to-point link of an elastic optical network using a reinforcement learning (RL) algorithm. We demonstrate a RL agent trained on idealized parameters in simulation, and compare the results with classical optimization algorithms such as a genetic algorithm and a simple power sweep. Furthermore, we present the high generalization ability of the RL agent under varying link conditions. Experimental results show that for 8 DWDM channels over nine 50km spans of True Wave Classic NZDSF with a span loss of 12 dB the RL agent is able to cope with these unseen conditions, removing the need to optimize for every possible link scenario during training.

A novel self-adaptive fractional multivariable grey model and its application in forecasting energy production and conversion of China

Energy production and conversion have a significant impact on the economic development of all countries in the world. China’s energy production and conversion are large. Therefore, accurate mid-to-long term China’s energy production and conversion forecasting is becoming more and more important for integrating energy systems and energy strategic planning. For this purpose, a novel fractional grey sequence is proposed based on Grunwald–Letnikov fractional calculus. Furthermore, a novel self-adaptive fractional multivariable grey model is proposed based on the novel sequence. In this article, we compare several classical optimization algorithms and finally choose Particle Swarm Optimization (PSO) to compute the parameters. In addition, Monte-Carlo simulation and probability density analysis (PDA) are presented in this article to verify the model’s performance. Monte-Carlo simulation reduces the randomness of the results of the model runs to a certain extent. Probability density analysis visualizes this randomness through kernel density estimation (KDE). This paper compares the new model with the existing seven grey models and predicts the total energy consumption per capita, energy conversion efficiency and total renewable energy in China, respectively. The experimental results show that the new model is superior to the other seven models in terms of stability and prediction accuracy.

Deep Learning LSTM Recurrent Neural Network Model for Prediction of Electric Vehicle Charging Demand

The immense growth and penetration of electric vehicles has become a major component of smart transport systems; thereby decreasing the greenhouse gas emissions that pollute the environment. With the increased volumes of electric vehicles (EV) in the past few years, the charging demand of these vehicles has also become an immediate requirement. Due to which, the prediction of the demand of electric vehicle charging is of key importance so that it minimizes the burden on the electric grids and also offers reduced costs of charging. In this research study, an attempt is made to develop a novel deep learning (DL)-based long-short term memory (LSTM) recurrent neural network predictor model to carry out the forecasting of electric vehicle charging demand. The parameters of the new deep long-short term memory (DLSTM) neural predictor model are tuned for its optimal values using the classic arithmetic optimization algorithm (AOA) and the input time series data are decomposed so as to maintain their features using the empirical mode decomposition (EMD). The novel EMD—AOA—DLSTM neural predictor modeled in this study overcomes the vanishing and exploding gradients of basic recurrent neural learning and is tested for its superiority on the EV charging dataset of Georgia Tech, Atlanta, USA. At the time of simulation, the best results of 97.14% prediction accuracy with a mean absolute error of 0.1083 and a root mean square error of 2.0628 × 10−5 are attained. Furthermore, the mean absolute error was evaluated to be 0.1083 and the mean square error pertaining to 4.25516 × 10−10. The results prove the efficacy of the prediction metrics computed with the novel deep learning LSTM neural predictor for the considered dataset in comparison with the previous techniques from existing works.

Efficient Parabolic Optimisation Algorithm for Adaptive VQE Implementations

Computational chemistry is one of the most promising applications of quantum computing, mostly thanks to the development of the Variational Quantum Eigensolver (VQE) algorithm. VQE is being studied extensively and numerous optimisations of VQE’s sub-processes have been suggested, including the encoding methods and the choice of excitations. Here we focus on a usually overlooked component of VQE, which is the choice of the classical optimisation algorithm. We introduce the parabolic optimiser that we designed specifically for the needs of VQE. We then continue to benchmark the parabolic optimiser against COBYLA for various implementations of VQE. We found that the parabolic optimizer consistently outperforms traditional optimization algorithms in multiple VQE implementations. Specifically, we show that the parabolic optimizer requires fewer quantum experiments compared to COBYLA while producing circuits with a similar amount of CNOTs.

Dual-Level Training of Gaussian Processes with Physically Inspired Priors for Geometry Optimizations.

Gaussian process (GP) regression has been recently developed as an effective method in molecular geometry optimization. The prior mean function is one of the crucial parts of the GP. We design and validate two types of physically inspired prior mean functions: force-field-based priors and posterior-type priors. In this work, we implement a dual-level training (DLT) optimizer for the posterior-type priors. The DLT optimizers can be considered as a class of optimization algorithms that belong to the delta-machine learning paradigm but with several major differences compared to the previously proposed algorithms in the same paradigm. In the first level of the DLT, we incorporate the classical mechanical descriptions of the equilibrium geometries into the prior function, which enhances the performance of the GP optimizer as compared to the one using a constant (or zero) prior. In the second level, we utilize the surrogate potential energy surfaces (PESs), which incorporate the physics learned in the first-level training, as the prior function to refine the model performance further. We find that the force-field-based priors and posterior-type priors reduce the overall optimization steps by a factor of 2-3 when compared to the limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) optimizer as well as the constant-prior GP optimizer proposed in previous works. We also demonstrate the potential of recovering the real PESs with GP with a force-field prior. This work shows the importance of including domain knowledge as an ingredient in the GP, which offers a potentially robust learning model for molecular geometry optimization and for exploring molecular PESs.

A modified random network distillation algorithm and its application in USVs naval battle simulation

Unmanned surface vessel (USV) operations will change the future form of maritime wars profoundly, and one of the critical factors for victory is the cluster intelligence of USVs. Training USVs for combat using reinforcement learning (RL) is an important research direction. Sparse reward as one of the complex problems in reinforcement learning causes sluggish and inefficient USV training. Therefore, a modified random network distillation (MRND) algorithm is proposed for the sparse reward problem. This algorithm measures the weight of internal rewards by calculating the variance of the number of training steps in each training episode to adjust internal and external rewards dynamically. Through the self-play iterative training method, our algorithm, in conjunction with the classical proximal policy optimization (PPO) algorithm, can improve USV cluster intelligence rapidly. Based on USV cluster combat training environments constructed on Unity3D and ML-Agent Toolkits platform, three types of USV cluster combat simulations are conducted to validate the algorithm, including a target pursuit combat simulation, a USV cluster maritime combat simulation, and a USV cluster base offense and defense combat simulation. Simulation experiments have shown that USV clusters trained with the MRND algorithm converge quicker, acquire more rewards in fewer steps, and exhibit a higher level of intelligence than the USV cluster trained by the comparison algorithms.

Multi-objective optimization of operation loop recommendation for kill web

In order to improve our military's level of intelligent accusation decision-making in future intelligent joint warfare, this paper studies operation loop recommendation methods for kill web based on the fundamental combat form of the future, i.e., “web-based kill,” and the operation loop theory. Firstly, we pioneer the operation loop recommendation problem with operation ring quality as the objective and closed-loop time as the constraint, and construct the corresponding planning model. Secondly, considering the case where there are multiple decision objectives for the combat ring recommendation problem, we propose for the first time a multi-objective optimization algorithm, the multi-objective ant colony evolutionary algorithm based on decomposition (MOACEA/D), which integrates the multi-objective evolutionary algorithm based on decomposition (MOEA/D) with the ant colony algorithm. The MOACEA/D can converge the optimal solutions of multiple single objectives non-dominated solution set for the multi-objective problem. Finally, compared with other classical multi-objective optimization algorithms, the MOACEA/D is superior to other algorithms superior in terms of the hyper volume (HV), which verifies the effectiveness of the method and greatly improves the quality and efficiency of commanders' decision-making.

A novel structure adaptive fractional discrete grey forecasting model and its application in China’s crude oil production prediction

Crude oil resources are related to all aspects of people's life, and play a vital role in the development of the national economy. Using nonlinear discrete data to reasonably predict crude oil production can help the government adjust energy structure and formulate energy development strategy, which has great practical significance. In this paper, a novel r-order accumulation operation with a parameter is proposed, and a novel structure adaptive fractional discrete grey forecasting model is established. Several classical optimization algorithms are compared, and the Grey Wolf Optimizer (GWO) is selected to calculate the parameters. For testifying the effectiveness of the model, a prediction model is constructed based on the total crude oil production in Qinghai, Liaoning and Shaanxi provinces of China, and a performance comparison experiment is designed with the existing six grey models. In addition, Monte Carlo simulation and probability density analysis provide a new perspective to further illustrate the robustness and accuracy of the proposed model. The experimental results show that this model is superior to the other six models in terms of fitting accuracy, prediction accuracy and model stability.

VQE Algorithm in Quantum Chemistry

For a complex quantum system, the dimensionality of the Hamiltonian grows exponentially with the size of the system, making it difficult to perform fast large-scale simulations and calculations on a classical computer. The first Variational Quantum Eigenvalue solver (VQE) proposed by Peruzzo et al. This gives the possibility of highly accurate molecular simulations and calculations in the field of quantum chemical computing. VQE is roughly based on the principle of using a quantum computer and set up a trial wave function |ψ(θ)⟩=U(θ)|ψ0⟩ to measure the energy expectation of the wave function |ψ(θ)⟩ at the corresponding initialization parameter, and then using a classical optimization algorithm to iterate θ0→θn over the energy to minimum until convergence. The significance of this paper is to verify the accuracy as well as the feasibility of the VQE algorithm for the calculation of LiH molecules in STO-3g basis, and to perform an error analysis and to verify the exponential acceleration of the classical algorithm implementation. These results shed light on guiding further exploration focusing on quantum computingt.

A Convolutional Neural Network-Based Surrogate Model for Multi-objective Optimization Evolutionary Algorithm Based on Decomposition

Evolutionary algorithms (EAs) show good performance in solving multi-objective optimization problems (MOPs). An EA needs to perform a substantial number of fitness evaluations. For the MOP with high complexity, the fitness evaluation functions are computationally expensive, making the evolutionary algorithms time-consuming. Surrogate-assisted evolutionary algorithms (SAEAs) that apply surrogate models instead of fitness exact evaluation functions have successfully reduced the computational complexity of fitness evaluations. However, because training a surrogate model requires a certain amount of calculation, a large amount of calculation is required by the SAEA to train multiple surrogate models. Furthermore, most existing surrogate models may not achieve desired evaluation accuracy when processing medium-dimensional and high-dimensional MOPs. This paper proposes a novel surrogate model. The surrogate model can be applied in multi-objective optimization evolutionary algorithm based on decomposition (MOEA/D), which is a classic decomposition-based multi-objective optimization algorithm. The surrogate model is designed based on the convolutional neural network structure, and it is called the multi-objective parallel fitness evaluation network (MPFEN). An MPFEN model contains multiple sub-networks which can be applied as the surrogate models. By training the MPFEN model, we can obtain all surrogate models required by a MOEA/D simultaneously without training each required surrogate model separately. Therefore, the amount of calculation of training surrogate models in a MOEA/D is reduced. The evaluation accuracy of the MPFEN model is tested by experiments. The experimental results show that the evaluation accuracy of MPFEN model is higher than that of other classical surrogate models in most cases. By applying the MPFEN model, the solution quality of SAEA is also improved.

An investigation of IBM quantum computing device performance on combinatorial optimisation problems

The intractability of deterministic solutions in solving \(\mathcal {NP}\)-Hard Combinatorial Optimisation Problems (COP) is well reported in the literature. One mechanism for overcoming this difficulty has been the use of efficient COP non-deterministic approaches. However, with the advent of quantum technology, these modern devices’ potential to overcome this tractability limitation requires exploration. This paper juxtaposes classical and quantum optimisation algorithms’ performance to solve two common COP, namely the Travelling Salesman Problem and the Quadratic Assignment Problem. Two accepted classical optimisation methods, Branch and Bound and Simulated Annealing, are compared to two quantum optimisation methods, Variational Quantum Eigensolver (VQE) algorithm and Quantum Approximate Optimisation Algorithm (QAOA). These algorithms are, respectively, executed on both classical devices and IBM’s suite of Noisy Intermediate-Scale Quantum (NISQ) devices. We have encoded the COP problems for the respective technologies and algorithms and provided the computational encodings for the NISQ devices. Our experimental results show that current classical devices significantly outperform the presently available NISQ devices, which both agree with and extend on those findings reported in the literature. Further, we introduce additional performance metrics to better compare the two approaches with respect to computational time, feasibility and solution quality. Our results show that the VQE performs better than QAOA with respect to these metrics, and we infer that this is due to the increased number of operations required. Additionally, we investigate the impact of a new set of basis gates on the quantum optimisation techniques and show they yield no notable improvement on obtained results. Finally, we highlight the present shortcomings of state-of-the-art NISQ IBM quantum devices and argue for continued future work on improving evolving devices.

Quantum Architecture Search with Meta‐Learning

AbstractVariational quantum algorithms (VQAs) have been successfully applied to quantum approximate optimization algorithms, variational quantum compiling and quantum machine learning models. The performances of VQAs largely depend on the architecture of parameterized quantum circuits (PQCs). Quantum architecture search (QAS) aims to automate the design of PQCs in different VQAs with classical optimization algorithms. However, current QAS algorithms do not use prior experiences and search the quantum architecture from scratch for each new task, which is inefficient and time consuming. In this paper, a meta quantum architecture search (MetaQAS) algorithm is proposed, which learns good initialization heuristics of the architecture (i.e., meta‐architecture), along with the meta‐parameters of quantum gates from a number of training tasks such that they can adapt to new tasks with fewer gradient updates, which leads to fast learning on new tasks. The proposed MetaQAS can be used with arbitrary gradient‐based QAS algorithms. Simulation results on variational quantum compiling (VQC) and quantum approximate optimization algorithm (QAOA) show that the architectures optimized by MetaQAS converge faster than a state‐of‐the‐art gradient‐based QAS algorithm, namely DQAS. MetaQAS also achieves a better solution than DQAS after fine‐tuning of gate parameters.

Implementable hybrid quantum ant colony optimization algorithm

We propose a new hybrid quantum algorithm based on the classical Ant Colony Optimization algorithm to produce approximate solutions for NP-hard problems, in particular optimization problems. First, we discuss some previously proposed Quantum Ant Colony Optimization algorithms, and based on them, we develop an improved algorithm that can be truly implemented on near-term quantum computers. Our iterative algorithm codifies only the information about the pheromones and the exploration parameter in the quantum state, while subrogating the calculation of the numerical result to a classical computer. A new guided exploration strategy is used in order to take advantage of the quantum computation power and generate new possible solutions as a superposition of states. This approach is specially useful to solve constrained optimization problems, where we can implement efficiently the exploration of new paths without having to check the correspondence of a path to a solution before the measurement of the state. As an example of a NP-hard problem, we choose to solve the Quadratic Assignment Problem. The benchmarks made by simulating the noiseless quantum circuit and the experiments made on IBM quantum computers show the validity of the algorithm.

CFHBA-PID Algorithm: Dual-Loop PID Balancing Robot Attitude Control Algorithm Based on Complementary Factor and Honey Badger Algorithm.

The PID control algorithm for balancing robot attitude control suffers from the problem of difficult parameter tuning. Previous studies have proposed using metaheuristic algorithms to tune the PID parameters. However, traditional metaheuristic algorithms are subject to the criticism of premature convergence and the possibility of falling into local optimum solutions. Therefore, the present paper proposes a CFHBA-PID algorithm for balancing robot Dual-loop PID attitude control based on Honey Badger Algorithm (HBA) and CF-ITAE. On the one hand, HBA maintains a sufficiently large population diversity throughout the search process and employs a dynamic search strategy for balanced exploration and exploitation, effectively avoiding the problems of classical intelligent optimization algorithms and serving as a global search. On the other hand, a novel complementary factor (CF) is proposed to complement integrated time absolute error (ITAE) with the overshoot amount, resulting in a new rectification indicator CF-ITAE, which balances the overshoot amount and the response time during parameter tuning. Using balancing robot as the experimental object, HBA-PID is compared with AOA-PID, WOA-PID, and PSO-PID, and the results demonstrate that HBA-PID outperforms the other three algorithms in terms of overshoot amount, stabilization time, ITAE, and convergence speed, proving that the algorithm combining HBA with PID is better than the existing mainstream algorithms. The comparative experiments using CF prove that CFHBA-PID is able to effectively control the overshoot amount in attitude control. In conclusion, the CFHBA-PID algorithm has great control and significant results when applied to the balancing robot.

Review of chaotic mapping enabled nature-inspired algorithms.

Chaotic maps were frequently introduced to generate random numbers and used to replace the pseudo-random numbers distributed in Gauss distribution in computer engineering. These improvements in optimization were called the chaotic improved optimization algorithm, most of them were reported better in literature. In this paper, we collected 19 classical maps which could all generate pseudo-random numbers in an interval between 0 and 1. Four types of chaotic improvement to original optimization algorithms were summarized and simulation experiments were carried out. The classical grey wolf optimization (GWO) and sine cosine (SC) algorithms were involved in these experiments. The final simulation results confirmed an uncertainty about the performance of improvements applied in different algorithms, different types of improvements, or benchmark functions. However, Results confirmed that Bernoulli map might be a better choice for most time. The code related to this paper is shared with https://gitee.com/lvqing323/chaotic-mapping.