Directed Particle Swarm Research Articles

Low-threshold dual-polarization electro-optic nonlinear activation functions

In this paper, a scheme for realizing low-threshold dual-polarization electro-optic nonlinear activation functions is proposed, designed, and experimentally demonstrated. This scheme relies on a structure composed of an optical power splitter and a thermo-optic modulator based on Mach–Zehnder interferometer, which are optimized by using direct binary search and particle swarm optimization algorithms. When a fixed proportion of the input optical signal is transformed into the electrical signal and the remaining input optical signal is modulated by using the thermo-optic effect, the rectified linear unit activation function with a low-threshold power of 0.2 mW for transverse-electric polarization and the Gaussian activation function with a low-threshold power of 0.1 mW for transverse-magnetic polarization can be measured separately. If the above-mentioned two nonlinear activation functions are introduced into the convolutional neural network to perform the modified National Institute of Standards and Technology handwritten digit classification task, validation accuracies of 97.3% and 96.85% will be achieved.

Particle Swarm Optimization for a redundant repairable machining system with working vacations and impatience in a multi-phase random environment

With the increasing reliance on cloud computing as the foundational manufacturing systems with intricate dynamics, featuring multiple service areas, varying job arrival rates, diverse service requirements, and the interplay of failures and impatience, significant analytical challenges arise. Queueing networks offer a powerful stochastic modeling framework to capture such complex dynamics. This paper develops a novel, exhaustive queueing model for a finite-capacity redundant multi-server system operating in a multi-phase random environment. The proposed model uniquely integrates real-world factors, including server breakdowns and repairs, waiting servers, synchronous working vacations, and state dependent balking and reneging, into a single queueing model, representing a significant advancement in the field. Using the matrix-analytic method, we establish the steady-state solution and derive key performance metrics. Numerical experiments and sensitivity analyses elucidate the impact of system parameters on performance measures. Additionally, a cost model is formulated, enabling cost optimization analysis using direct search method and Particle Swarm Optimization (PSO) to identify efficient operating configurations.

Cost optimization and reliability analysis of fault tolerant system with service interruption and reboot

Due to widespread usage in many real time systems, reliability modeling and cost optimization of fault tolerance system have drawn attention of the practitioners. The fault tolerance in these systems can be provided by the support of maintenance and redundant components that help in smooth operation of the system in spite of failure of some active components. This investigation deals with the performance modeling of a fault-tolerant system consisting of a finite number of active (online) and standby components. During the switching from active to standby, the recovery procedure is performed, which may be imperfect. In case of imperfect recovery, the system reboot takes place. The maintenance of all the components is managed by a repairman (server) which is subject to failure. When the server is interrupted for rendering the service, functioning does not get stopped due to the system switch-over from perfect working to working breakdown mode. The system works even when the server is on working vacation and performs repair jobs of the failed components. The machine repair model based on Markovian process is developed to derive the transient probabilities and other performance indices of the fault tolerant system using Laplace transforms and matrix analytical method. Using the direct search strategy and particle swarm optimization, the cost-benefit analysis is done. The optimal design of the control parameters for the fault-tolerant system are presented by framing a cost-effective ratio function. The model is examined computationally by performing the numerical simulation and cost optimization.

One-step vs horizon-step training strategies for multi-step traffic flow forecasting with direct particle swarm optimization grid search support vector regression and long short-term memory

In the increasingly complex urban transportation landscape, the necessity for accurate, multi-step forecasting has never been more apparent to help plan for long-term, strategic transportation initiatives like Intelligent Transportation Systems. This work focuses on resolving uncertainties around the proper training step size for machine learning models performing multi-step traffic flow forecasting. Two strategies are considered: one-step and horizon-step training sizes, which extend from the single-step forecasting method. This paper compares and evaluates two machine learning models: the Support Vector Regression and Long Short-Term Memory (LSTM). Data from two country road locations, including urban roads in Kuala Lumpur and the I5-North freeway in California, are employed to forecast traffic flow in 1-hour increments, projecting from 2 h up to 24 h ahead. The results reveal a significant difference in performance between the two training step sizes, with the one-step training size emerging as the more consistent and optimal strategy for both Direct and Multi-Input Multi-Output forecasting strategies. The proposed model, the Direct Particle Swarm Optimization Grid Search Support Vector Regression (Direct-PSOGS-SVR) model, exhibited comparable or slightly better performance than the LSTM model in both environments. In the I5-North freeway dataset, the Direct-PSOGS-SVR achieved similar Root Mean Squared Error values across most forecasting tasks compared to the LSTM, indicating its effectiveness in stable traffic flow patterns. Notably, in the more dynamic urban environment of Kuala Lumpur, the Direct-PSOGS-SVR model also demonstrated stability and resilience in forecasting accuracy, effectively handling the inherent noise and fluctuations in traffic patterns. These findings serve as a valuable guide for practitioners in selecting the most efficacious combination of training strategies for specific time series forecasting tasks utilizing machine learning models, particularly in traffic flow forecasting. Introducing the Direct-PSOGS-SVR model enriches the landscape of machine learning solutions for traffic forecasting, underscoring the paper’s contributions to the broader understanding of time series forecasting dynamics.

Optimized design of nano-photonic devices for temperature self-regulating on vanadium dioxide thin films

Phase change materials can enable temperature self-regulation due to their drastic changes in optical properties accompanying the phase transition. Significant reduction of the optical absorption after the transition is the key ingredient for an enhanced regulating performance. However, the absorptivity of unpatterned vanadium dioxide (VO2) thin films can hardly be reduced after phase transition at visual-to-infrared band. In this work, we combine the direct binary search (DBS) and particle swarm optimization (PSO) algorithms for an optimized design of temperature self-regulating nano-photonic devices on VO2 thin films. For a given incident wavelength, a pixelated structure is firstly inverse-designed by the DBS algorithm which maximizes the absorption contrast before and after the transition. To overcome fabrication challenges as pixel size is at deep sub-wavelength scale, the pixelated structure can then be replaced by geometric shapes which are more tractable in manufacturing processes. The geometrical parameters are optimized by the PSO algorithm where our optimized device brings the absorptivity down to 33% after the transition. These results provide an effective way for the inverse design of optimized nano-photonic structures based on phase change materials.

Particle swarm optimization-assisted approach for the extraction of VCSEL model parameters.

We propose a direct particle swarm optimization (PSO) method for extracting the parameters of a physical model describing the behavior of vertical-cavity surface-emitting lasers (VCSELs), starting from the light-current (L-I) characteristics and the small signal modulation (S21) responses, at different currents and temperatures. With an optimal choice of hyperparameters of the algorithm, the method is able to predict parameters that accurately reproduce the behavior of the device. Its prediction capabilities are compared to those of two commonly used nonlinear optimizers (Interior Point and Levenberg-Marquardt), to benchmark its performances.

Elite Directed Particle Swarm Optimization with Historical Information for High-Dimensional Problems

High-dimensional optimization problems are ubiquitous in every field nowadays, which seriously challenge the optimization ability of existing optimizers. To solve this kind of optimization problems effectively, this paper proposes an elite-directed particle swarm optimization (EDPSO) with historical information to explore and exploit the high-dimensional solution space efficiently. Specifically, in EDPSO, the swarm is first separated into two exclusive sets based on the Pareto principle (80-20 rule), namely the elite set containing the top best 20% of particles and the non-elite set consisting of the remaining 80% of particles. Then, the non-elite set is further separated into two layers with the same size from the best to the worst. As a result, the swarm is divided into three layers. Subsequently, particles in the third layer learn from those in the first two layers, while particles in the second layer learn from those in the first layer, on the condition that particles in the first layer remain unchanged. In this way, the learning effectiveness and the learning diversity of particles could be largely promoted. To further enhance the learning diversity of particles, we maintain an additional archive to store obsolete elites, and use the predominant elites in the archive along with particles in the first two layers to direct the update of particles in the third layer. With these two mechanisms, the proposed EDPSO is expected to compromise search intensification and diversification well at the swarm level and the particle level, to explore and exploit the solution space. Extensive experiments are conducted on the widely used CEC’2010 and CEC’2013 high-dimensional benchmark problem sets to validate the effectiveness of the proposed EDPSO. Compared with several state-of-the-art large-scale algorithms, EDPSO is demonstrated to achieve highly competitive or even much better performance in tackling high-dimensional problems.

Direct Torque Control of Induction Motor Based on Particle Swarm Optimization

Induction Motor (IM) is most commonly used in different industrial applications, that require fast dynamic response and accurate control over wide speed ranges. Therefore, this paper proposes high performance Direct Torque Control (DTC), which is closed loop control system to improve the dynamic performance of IM. The main objective of this work is to improve the speed response of IM during different load and speed conditions. Proportional-Integral (PI) controller based on Particle Swarm Optimization (PSO) technique is used for optimal gains tuning. The results show the improvement in the speed response of DTC using PSO technique when compared with conventional PI and the conventional DTC (without PI), in terms of reducing steady state error, settling time, overshoot and ripple reduction which make this controller more robust to variation in load and speed. The simulation of the overall drive system is performed using MATLAB/Simulink program version 7.10 (R2010a).Keywords: Induction Motor (IM), Direct torque control, PI controller and Particle Swarm Optimization (PSO).

A New Nanorobots Movement Control Strategy for Treating Cancer

Nanorobots were proposed to deliver drugs directly into cancer cells to destroy only these cells without harming the surrounding cells. During their journey, the nanorobots may encounter some obstacles such as blood cells which may be resistant to their movement. So, it is necessary to avoid collisions with these obstacles to achieve their goal. This study proposes a new strategy for controlling the nanorobots movement in human body to reach cancer cells. This proposed strategy uses an efficient algorithm based on fuzzy logic for dynamic obstacle avoidance. Also, this proposed strategy uses the directed particle swarm optimization (DPSO) algorithm for delivering nanorobots to cancer cells. Simulation experiments have proved that the proposed control strategy can efficiently deliver nanorobots to their target and also avoid collisions with dynamic obstacles which move in the same direction of the nanorobots or across their direction.

Directed particle swarm optimization with Gaussian-process-based function forecasting

Particle swarm optimization (PSO) is an iterative search method that moves a set of candidate solution around a search-space towards the best known global and local solutions with randomized step lengths. PSO frequently accelerates optimization in practical applications, where gradients are not available and function evaluations expensive. Yet the traditional PSO algorithm ignores the potential knowledge that could have been gained of the objective function from the observations by individual particles. Hence, we draw upon concepts from Bayesian optimization and introduce a stochastic surrogate model of the objective function. That is, we fit a Gaussian process to past evaluations of the objective function, forecast its shape and then adapt the particle movements based on it. Our computational experiments demonstrate that baseline implementations of PSO (i. e., SPSO2011) are outperformed. Furthermore, compared to, state-of-art surrogate-assisted evolutionary algorithms, we achieve substantial performance improvements on several popular benchmark functions. Overall, we find that our algorithm attains desirable properties for exploratory and exploitative behavior.

Directed jaya algorithm for delivering nano-robots to cancer area

In the last few years, it was proposed to deliver drugs using Nano-robots for treating cancer. This paper compares between two recent and efficient algorithms for delivering Nano-robots to cancer area. These algorithms are Jaya algorithm and Directed Particle Swarm Optimization (DPSO) algorithm. In this paper, we also propose a new hybrid algorithm that combines Jaya and DPSO to speed up the process of Nano-robots delivery. The proposed algorithm is called Directed Jaya (DJaya) algorithm. Experiments have proved that the efficiency of DJaya is higher than both Jaya and DPSO. We show experimentally that DJaya starts delivering Nano-robots earlier than DPSO to facilitate the initiation of the drug release. Also, DJaya finishes delivering Nano-robots earlier than Jaya to complete the drug dose. In addition to this, DJaya groups the Nano-robots together in the target area like DPSO to speed up the drug release process. We finally propose a new strategy for destroying cancer cells efficiently with relatively small number of Nano-robots. This strategy can save 40% of Nano-robots.

Indoor localization of wireless emitter using direct position determination and particle swarm optimization

Indoor localization of wireless emitter using direct position determination and particle swarm optimization

Plug-in direct particle swarm repetitive controller with a reduced dimensionality of a fitness landscape – a multi-swarm approach

Abstract The paper describes a modification to the recently developed plug-in direct particle swarm repetitive controller (PDPSRC) for the sine-wave constant-amplitude constant-frequency (CACF) voltage-source inverter (VSI). The original PDPSRC algorithm assumes that the particle swarm optimizer (PSO) takes into account a performance index defined over the whole reference signal period. Each particle stores all the samples of the control signal, e.g. α = 200 samples for a controller working at 10 kHz and the reference frequency equal to 50 Hz. Therefore, the fitness landscape (i.e. the performance index) is -dimensional ( D), which makes optimization challenging. That solution can be categorized as the single-swarm one. It has been previously shown that the swarm controller does not suffer from long-term stability issues encountered in the classic iterative learning controllers (ILC). However, the convergence of the swarm has to be kept at a relatively low rate to enable successful exploitation in the D search space, which in turn results in slow responsiveness of the PDPSRC. Here a multi-swarm approach is proposed in which we divide a dynamic optimization problem (DOP) among less dimensional swarms. The reference signal period is segmented into shorter intervals and the control signal is optimized in each interval independently by separate swarms. The effectiveness of the proposed approach is illustrated with the help of numerical experiments on the CACF VSI with an output LC filter operating under nonlinear loads.

Design of Robust Fractional PID Controller Using Triangular Strip Operational Matrices

Abstract The present article proposes a simple tuning technique aimed to produce a robust non-integer order PID controller exhibiting iso-damping property during reparameterization of a plant. The required robustness property is achieved by letting the fractional PID control system to imitate the dynamics of a reference system having Bode’s ideal transfer function in its forward path. The objective of designing robust controller by tracking the dynamics of reference control system is defined mathematically as an H∞- optimal control problem. Fractional differential systems are transformed into algebraic equations by the use of triangular strip operational matrices. The H∞-optimal control problem is then changed to an ∞-norm minimization of a parameter (Kc,KI,Kd, λ, μ) varying square matrix. Global optimization techniques, Luus-Jaakola direct search and particle swarm optimization are employed to find the optimum values of fractional PID controller parameters. The proposed method of control system design is implemented in heating furnace temperature control, automatic voltage regulator system and some integer and fractional order process models. Fractional PI, fractional PD and various versions of fractional PID controllers are designed as optimal controllers using the triangular strip operational matrix based control design method.

A Performance Study on Synchronous and Asynchronous Update Rules for A Plug-In Direct Particle Swarm Repetitive Controller

Abstract In this paper two different update schemes for the recently developed plug-in direct particle swarm repetitive controller (PDPSRC) are investigated and compared. The proposed approach employs the particle swarm optimizer (PSO) to solve in on-line mode a dynamic optimization problem (DOP) related to the control task in the constant-amplitude constant-frequency voltage-source inverter (CACF VSI) with an LC output filter. The effectiveness of synchronous and asynchronous update rules, both commonly used in static optimization problems (SOPs), is assessed and compared in the case of PDPSRC. The performance of the controller, when synthesized using each of the update schemes, is studied numerically.

A Novel Intrusion Detection Approach using Multi-Kernel Functions

Network intrusion detection finds variant applications in computer and network industry. How to achieve high intrusion detection accuracy and speed is still received considerable attentions in this field. To address this issue, this work presents a novel method that takes advantages of multi-kernel computation technique to realize speedy and precise network intrusion detection and isolation. In this new development the multi-kernel function based kernel direct discriminant analysis (MKDDA) and quantum particle swarm optimization (QPSO) optimized kernel extreme learning machine (KELM) were appropriately integrated and thus form a novel method with strong intrusion detection ability. The MKDDA herein was firstly employed to extract distinct features by projecting the original high dimensionality of the intrusion features into a low dimensionality space. A few distinct and efficient features were then selected out from the low dimensionality space. Secondly, the KELM was proposed to provide quick and accurate intrusion recognition on the extracted features. The only parameter need be determined in KELM is the neuron number of hidden layer. Literature review indicates that very limited work has addressed the optimization of this parameter. Hence, the QPSO was used for the first time to optimize the KELM parameter in this paper. Lastly, experiments have been implemented to verify the performance of the proposed method. The test results indicate that the proposed LLE-PSO-KELM method outperforms its rivals in terms of both recognition accuracy and speed. Thus, the proposed intrusion detection method has great practical importance.

A plug-in direct particle swarm repetitive controller for a single-phase inverter

The paper presents an online particle swarm optimizer (PSO) as an iterative learning controller for the single phase inverter with an output LC filter. The novelty of the solution lies in the fact that the swarm directly stores samples of the control signal. The swarm optimizes, according to a user-defined performance index, in online mode the control signal to reject the repetitive disturbance (the load current drawn, for example, by the diode rectifier). The concept of the direct swarm controller is investigated with the help of numerical simulations. Streszczenie. W artykule opisano regulator rojowy realizuj ˛ acy sterowanie z uczeniem iteracyjnym dla jednofazowego falownika napi ˛ ecia z wyj´ filtrem LC. Oryginalno´ s´ c rozwi ˛ azania polega na fakcie bezpoprzechowywania probek sygnalu steruj ˛ acego przez roj cz ˛ astek. Roj optymalizuje w trybie on-line sygnal steruj ˛ acy eliminuj ˛ ac wplyw okresowego zaklocenia (pr ˛ adu obci ˛ a˙ zenia pobieranego, na przyklad, przez prostownik diodowy z filtrem pojemno´ sciowym) na jako´ snapi ˛ ecia wyj´ sciowego. Sygnal steruj ˛ acy jest optymalny z uwagi na zdefiniowany przez uwskaznik jakoKoncepcja bezpo´ sredniego regulatora rojowego zostala zbadana przy utechnik modelowania numerycznego. (Sterowanie powtarzalne napi ˛ eciem wyj ´ sciowym falownika przy uzyciu bezporegulatora rojowego)

Leapfrogging and synoptic Leapfrogging: A new optimization approach

A novel optimization technique is introduced and demonstrated. Leapfrogging starts with a randomly located set of trial solutions (termed players) within the feasible decision variable (DV) space. At each iteration, the player with the worst objective function (OF) value is relocated to a random position within its DV-space reflection on the other side of the player with the best OF value. Test cases reveal that this simple algorithm has benefits over classic direct and gradient-based methods and particle swarm in speed of finding the optimum and in handling surface aberrations, including ridges, multi-optima, and stochastic objective functions. Potential limitations and analysis opportunities are discussed.

Improved Dynamic K-Coverage Algorithms in Mobile Sensor Networks

In this paper, four PSO based distributed algorithms are presented to attain k-coverage in the target filed. In the first algorithm named K-Coverage Particle Swarm Optimization (KPSO), each static sensor which discovers an event in its sensing range, implements Particle Swarm Optimization (PSO) algorithm in a distributed manner on its mobile sensors. The calculation time is considered as a big bottleneck in PSO, so a second algorithm named K-Coverage Virtual Force directed Particle Swarm Optimization (KVFPSO) is presented, comprised of Virtual Force and KPSO algorithms. In the first and second proposed algorithms, the best experiences of the particles were used to determine their speed. It is possible that these responses might not be the final result and cause extra movements. Another algorithm named KVFPSO-Learning Automata (KVFPSO-LA) is introduced based on which the speed of particles is corrected by using the existing knowledge and the feedback from the actual implementation of the algorithm. To improve performance of the algorithm, Improved KVFPSO-LA is introduced, in which static sensors are equipped with learning automata. Simulation results show that the proposed protocols perform well with respect to balanced energy consumption among nodes, thus maximizing network life-time.

Kinematic Analysis and Optimization of a New Compliant Parallel Micromanipulator

In this paper, a new three translational degrees of freedom (DOF) compliant parallel micromanipulator (CPM) is proposed, which has an excellent accuracy of parallel mechanisms with flexure hinges. The system is established by a proper selection of hardware and analyzed via the derived pseudo-rigid-body model. In view of the physical constraints imposed by both the piezoelectric actuators and flexure hinges, the CPM's reachable workspace is determined analytically, where a maximum cylinder defined as an usable workspace can be inscribed. Moreover, the optimal design of the CPM with the consideration of the usable workspace size and global dexterity index simultaneously is carried out by utilizing the approaches of direct search method, genetic algorithm (GA), and particle swarm optimization (PSO), respectively. The simulation results show that the PSO is the best method for the optimization, and the results are valuable in the design of a new micromanipulator.