Dynamic Differential Evolution Algorithm Research Articles

Dynamic multi-strategy integrated differential evolution algorithm based on reinforcement learning for optimization problems

The introduction of a multi-population structure in differential evolution (DE) algorithm has been proven to be an effective way to achieve algorithm adaptation and multi-strategy integration. However, in existing studies, the mutation strategy selection of each subpopulation during execution is fixed, resulting in poor self-adaptation of subpopulations. To solve this problem, a dynamic multi-strategy integrated differential evolution algorithm based on reinforcement learning (RLDMDE) is proposed in this paper. By employing reinforcement learning, each subpopulation can adaptively select the mutation strategy according to the current environmental state (population diversity). Based on the population state, this paper proposes an individual dynamic migration strategy to “reward” or “punish” the population to avoid wasting individual computing resources. Furthermore, this paper applies two methods of good point set and random opposition-based learning (ROBL) in the population initialization stage to improve the quality of the initial solutions. Finally, to evaluate the performance of the RLDMDE algorithm, this paper selects two benchmark function sets, CEC2013 and CEC2017, and six engineering design problems for testing. The results demonstrate that the RLDMDE algorithm has good performance and strong competitiveness in solving optimization problems.

Energy-Saving Optimization Method of Urban Rail Transit Based on Improved Differential Evolution Algorithm

The transformation of railway infrastructure and traction equipment is an ideal way to realize energy savings of urban rail transit trains. However, upgrading railway infrastructure and traction equipment is a high investment and difficult process. To produce energy-savings in the urban rail transit system without changing the existing infrastructure, we propose an energy-saving optimization method by optimizing the traction curve of the train. Firstly, after analyzing the relationship between the idle distance and running energy-savings, an optimization method of traction energy-savings based on the combination of the inertia motion and energy optimization is established by taking the maximum idle distance as the objective; and the maximum allowable running speed, passenger comfort, train timetable, maximum allowable acceleration and kinematics equation as constraints. Secondly, a solution method based on the combination of the adaptive dynamic multimodal differential evolution algorithm and the Q learning algorithm is applied to solve the optimization model of energy-savings. Finally, numeric experiments are conducted to verify the proposed method. Extensive experiments demonstrate the effectiveness of the proposed method. The results show that the method has significant energy-saving properties, saving energy by about 11.2%.

Photovoltaic parameters estimation of poly-crystalline and mono-crystalline modules using an improved population dynamic differential evolution algorithm

<span>Photovoltaic (PV) parameters estimation from the experimental current and voltage data of PV modules is vital for monitoring and evaluating the performance of PV power generation systems. Moreover, the PV parameters can be used to predict current-voltage (I-V) behavior to control the power output of the PV modules. This paper aimed to propose an improved differential evolution (DE) integrated with a dynamic population sizing strategy to estimate the PV module model parameters accurately. This study used two popular PV module technologies, i.e., poly-crystalline and mono-crystalline. The optimized PV parameters were validated with the measured data and compared with other recent meta-heuristic algorithms. The proposed population dynamic differential evolution (PDDE) algorithm demonstrated high accuracy in estimating PV parameters and provided perfect approximations of the measured I-V and power-voltage (P-V) data from real PV modules. The PDDE obtained the best and the mean RMSE value of 2.4251E-03 on the poly-crystalline Photowatt-PWP201, while the best and the mean RMSE value on the mono-crystalline STM6-40/36 was 1.7298E-03. The PDDE algorithm showed outstanding accuracy performance and was competitive with the conventional DE and the existing algorithms in the literature.</span>

Dynamic hybrid mechanism-based differential evolution algorithm and its application

In order to effectively schedule railway train delay, an adaptive cooperative co-evolutionary differential evolution with dynamic hybrid mechanism of the quantum evolutionary algorithm and genetic algorithm, named QGDECC is designed in this paper. In the QGDECC, the quantum variable decomposition strategy is designed by utilizing qubit string to decompose variables adaptively according to the coevolution performance. Then the increment mutation method is proposed to improve the convergence speed which make full use of searched evolution information. Besides, the parameter adaptive strategy is deeply explored for strengthening the robust of the algorithm. The QGDECC with global search capability is employed to realize a railway train delay scheduling method for effectively eliminating the impact of train delay. Finally, several benchmark functions and actual train operation data are selected to verify the optimization performance of QGDECC. The experimental results show that QGDECC has higher adaptability, faster convergence speed and accuracy. The train delay scheduling method can effectively eliminate the impact of delay on the railway network, and minimize the gap between the rescheduled train schedule and the original train schedule.

Comparison of coal-to-ethanol product separation strategies

Given China's energy structure and the limitations of bioethanol, the coal-to-ethanol (CTE) pathway, from dimethyl ether to ethanol (DMTE) via carbonylation and hydrogenation, is highly anticipated. Ethanol, methanol, methyl acetate, and ethyl acetate are the crude hydrogenation products that need to be purified, requiring at least an eight-column scheme. However, the optimization of the existing separation strategy with ethanol as the priority is unfavorable in the following aspects: it is usually plagued by tedious rules of thumb and, due to the large scale of the process, is prone to falling into local minima; pre-designed heat integration inevitably neglects the interaction of parameter optimization and heat integration; reports on alternative feasible distillation sequences are scarce in publications, let alone comparisons amongst these counterparts. Therefore, four viable separation strategies are proposed in this paper to compare with this faulted separation strategy. A self-adapting dynamic differential evolution (SADDE) algorithm, which is accelerated by parallel computation, is used to search for optimal column parameters of all the configuration options and facilitates simultaneous heat integration structure synthesis. Two strategies stand out after 3000 generations of evolution. Splitting methanol outperforms in specific steam consumption (SSC) of ethanol (1.8177), much better than the benchmark (2.4840), and splitting ethyl acetate with ethyl acetate priority has the most competitive total annual cost (TAC), 23.98% lower than the benchmark.In summary, this paper provides a reference for optimizing complex distillation systems like CTE product separation, or more specifically, the DMTE route, before the appearance of the most suitable separation strategy in demand. Furthermore, it will also serve for the CTE superstructure to further explore the optimal distillation sequence.

A dynamic population reduction differential evolution algorithm combining linear and nonlinear strategy piecewise functions

AbstractThe population size has a great impact on the performance of the Differential Evolution (DE) algorithm, but in many classic DE algorithms, the population size is usually determined by the user based on the experience value, and it remains unchanged during the evolution process, which greatly affect the performance of DE. To this end, a dynamic population reduction differential evolution algorithm (DPSHADE) that combines linear and nonlinear strategy piecewise functions is proposed. The algorithm uses a dynamic population size reduction method to dynamically adjust the population size during operation, and construct a combination of linear and nonlinear piecewise functions for dynamic scale adaptive adjustment. In this paper, we proposed the DPSHADE algorithm and is compared with the four traditional algorithms in the CEC2017 benchmark set. The experimental results show that DPSHADE performs better in overall performance, which is significant better than the performance of SHADE.

Dynamic multi-objective differential evolution algorithm based on the information of evolution progress

Dynamic multi-objective differential evolution algorithm based on the information of evolution progress

Development of Admittance Control Method with Parameter Self-optimization for Hydraulic Series Elastic Actuator

Series elastic actuators (SEAs) have emerged as promising devices to enhance the safety of human-robot interactions in the manufacturing industry. However, the control of a hydraulic SEA under disturbance remains an unexplored issue. To address this problem, an admittance control method with parameter self-optimization is developed in this study. The hydraulic SEA and its dynamic model are first developed, and then, an admittance controller that combines a passive disturbance observer (DOB) and a feedback compensator is developed based on load movement dynamics. The control law of the framework is made independent of the hydraulic dynamics by considering the uncertainty and tracking error as disturbances. This simplifies the controller computation, enhances system robustness, and facilitates practical application. Next, the control performance is further improved by optimizing the control parameters using an improved crowding-based dynamic population size differential evolution (crowding-based dynNP-DE) algorithm. Benchmark and optimization experiments are performed to verify the superiority of the modified algorithm and obtain the control parameters. Finally, the optimized parameters are applied to practical experiments to validate the improved performance of the proposed admittance control scheme. The results show that the proposed method effectively reduces the SEA stiffness tracking error, with respect to the external contact force.

A self-adaptive multi-objective dynamic differential evolution algorithm and its application in chemical engineering

This paper proposes a new multi-objective dynamic differential evolution algorithm with parameter self-adaptive strategies, named SA-MODDE. All components of the algorithm are synergically designed to reach its full potential, containing parental selection, mutation strategy, parameter setting, survival selection, constraint handling, and termination criteria. The improvement measures emphasize exploiting Pareto dominance information more efficiently. Particularly, parameter adaptation schemes are introduced based on both prior knowledges of current individual and feedback information on previous promising solutions, and their effectiveness is validated by comparison with three fixed-parameter combinations. Extensive numerical tests are conducted on multiple test suites with five state-of-the-art peer competitors. The statistical results demonstrated that the SA-MODDE exhibits good proximity and diversity in dealing with benchmark functions with various characteristics. Three industrial (bio)chemical processes, including two optimal control and one reformulated constrained tri-objective, are investigated to show the feasibility and robustness of the SA-MODDE.

Dynamic Security and Stability Region Under Different Renewable Energy Permeability in IENGS System

The integration of renewable energy sources (RESs) remains growing rapidly in recent years. At the same time, security problems after contingencies were attracting great attention, due to deficiencies for the stable operation, even could induce the collapse of integrated energy systems with the deep interconnection of such RESs. To existing issues, the hyper-plane (HP) expression is proved to be an excellent tool for power system situational awareness and stability-constrained operation in practical dynamic security stability regions (PDSSR). In this study, a concept and model were proposed for a high renewable energy penetration of integrated electricity-natural gas system (HRE-IENGS) based on the N - 1 security guideline. The scheme instruction and system evaluation processes as follows: in order to improve the accuracy of the IENGS system, simulation fitting solutions of PDSSR boundary were presented according to a hybrid algorithm, including dynamic self-adaptive differential evolution algorithm (DSADE), and a least square support vector machine (LSSVM) algorithm. The adopted algorithm approach was applied to rapid approximating of the PDSSR boundary by HP expression in power injection spaces. Simultaneously, different renewable energy permeabilities (DREPs) were set from 5 to 50 %, and active power injections of DREPs under short circuit fault on IEEE 118 bus system simulated by using DIgSLENT/power factory software. As the consequence, this mentioned method effectively described the security stability boundary of HRE-IENGS, and realized the three-dimensional visualization space of DREPs-PDSSR. In addition, the behavior of operation points and PDSSR under DREPs were carefully investigated, and unstable runs were accurately exposed. Through the PDSSR, the state analysis could be conducted rapidly on several parameters, including security and stability assessment with various energy supply capabilities. Meanwhile, these indexes were calculated offline and applied on-line. The results of this study verified the accuracy and effectiveness of the proposed modeling for the considered system, and thus could provide technical support for the stability of the HRE-IENGS system.

Proton exchange membrane fuel cell model parameter identification based on dynamic differential evolution with collective guidance factor algorithm

This paper firstly proposes a dynamic differential evolution algorithm (DDE-CGF) with a collective guiding factor to solve the problem of parameter identification and optimization of proton exchange membrane fuel cell (PEMFC) model. Inspired by the swarm intelligence scheme, a collective guidance factor is designed to accelerate the convergence speed without affecting the convergence accuracy. Moreover, a dynamic scaling factor and the dynamic crossover probability based on evolutionary mechanism are introduced to enhance the diversity of population as well as improve the global searching performance. Through testing eight benchmark functions, the DDE-CGF algorithm exhibits superior performance in both convergence accuracy and speed. Based on the excellent global performance, applying DDE-CGF algorithm to the parameter identification of the PEMFC model, and more accurate parameter values are obtained. Comparing with other algorithms, the result proves that the DDE-CGF algorithm could accurately estimate model parameters and the identified model could greatly describe the dynamical characteristic of the PEMFC model.

Adaptive weighted dynamic differential evolution algorithm for emergency material allocation and scheduling

AbstractEmergency material allocation and scheduling is a combination optimization problem, which is essentially a Non‐deterministic Polynomial (NP) problem. Aiming at the problems such as slow convergence, easy prematurely falling into local optimum, and parameter constraints to solve high‐dimensional and multi‐modal combination optimization problems, this article proposes an adaptive weighted dynamic differential evolution (AWDDE) algorithm. The algorithm uses a chaotic mapping strategy to initialize the population. By weighting the standard differential evolution (DE) mutation strategy, a new weighted mutation operator is proposed. The scaling factor and cross probability can be adaptively adjusted. A disturbance operator is introduced to randomly generate the perturbation mutation and to accelerate the premature individuals to jump out of the local optimum. The algorithm is applied to the problem of emergency material allocation and scheduling, and a two‐stage emergency material allocation and scheduling model is established. Compared with the standard DE algorithm and the chaos adaptive particle swarm algorithm, the results show that the AWDDE algorithm has the characteristics of stronger global optimization ability and faster convergence speed compared with other optimization algorithms, which provide assistance for smart cities research, including smart city services, applications, case studies, and policymaking considerations for emergency management.

Synthesis of Sparse Rectangular Planar Arrays with Multiple Constraints Based on Dynamic Parameters Differential Evolution Algorithm

For solving the problem of the synthesis of sparse rectangular planar arrays with multiple constraints, this paper proposes a Dynamic Parameters Differential Evolution (DPDE) based algorithm. Firstly, to improve searching efficiency and accuracy of Differential Evolution (DE), the proposed method introduces dynamically changing strategies to the scaling factor and the crossover probability of the traditional Differential Evolution algorithm. Secondly, a modified matrix mapping method and the redefinition of mapping principles are presented to make up the defects of strong randomness and low accuracy in existing methods. Finally, simulation experiments of antenna arrays are performed to validate the effectiveness of the proposed method, and the results demonstrate that the proposed method performs out the existing methods in the respect of reducing peak sidelobe level of antenna arrays.

Metaheuristics for two-stage flow-shop assembly problem with a truncation learning function

ABSTRACT This study examines a two-stage three-machine flow-shop assembly scheduling model in which job processing time is considered as a mixed function of a controlled truncation parameter with a sum-of-processing-times-based learning effect. However, the truncation function is very limited in the two-stage flow-shop assembly scheduling settings. To overcome this limitation, this study investigates a two-stage three-machine flow-shop assembly problem with a truncation learning function where the makespan criterion (completion of the last job) is minimized. Given that the proposed model is NP hard, dominance rules, lemmas and a lower bound are derived and applied to the branch-and-bound method. A dynamic differential evolution algorithm, a hybrid greedy iterated algorithm and a genetic algorithm are also proposed for searching approximate solutions. Results obtained from test experiments validate the performance of all the proposed algorithms.

A material handling scheduling method for mixed-model automotive assembly lines based on an improved static kitting strategy

Since the diversification of customer demands poses a great challenge for manufacturing enterprises and the scheduling problem of material handling affects the efficiency of assembly lines, this paper proposes a novel scheduling method, an improved static kitting strategy, to solve the scheduling problems of the material handling for automotive mixed-model assembly lines (MMALs) based on line-integrated supermarkets. Firstly, an integer programming mathematical model is established with the objective of minimizing the number of logistic workers. Then, an improved static kitting strategy is presented to solve the problem and a model based on graph theories is constructed to transform the scheduling problem to a mathematical one. Afterwards, a Kuhn-Munkres algorithm and an elite opposition-based learning adaptive dynamic differential evolution algorithm, named EOADDE algorithm, is developed to solve the scheduling problem. The elite opposition-based learning (EOL) and self-adaptive operators are applied to the proposed EOADDE algorithm to enhance the local search ability and the convergence speed. Finally, computational experiments of the proposed algorithm are carried out compared with benchmark algorithms, and the feasibility and effectiveness of proposed methods are verified by results.

Stochastic optimal operation of flexible distribution networks for security improvement considering active management

This paper presents a stochastic optimal operation problem of gas turbine integrated distribution networks in the presence of active management schemes, which is formulated as a multi-objective chance-constrained mixed integer nonlinear programming problem. The control variables are the on-load tap-changer tap position, the power provided by the distributed generation (DG), the DG power factor angle, the load participating in demand side management and the switch status. The objectives defined in this paper are to simultaneously minimize the expectation cost and variation coefficient of security distance. Uncertainties related to DG output and load fluctuation and fault power restoration under contingencies are also considered in the optimization problem. The collaboration of normal boundary intersection and dynamic niche differential evolution algorithm is proposed to handle the optimal operation mode. Simulation results are presented and demonstrate the effectiveness of the proposed model. Compared with the operation result without the consideration of security, the security-constrained operation can reduce the expectation cost. So the proposed optimization is reasonable and valuable.

Optimized railway track health monitoring system based on dynamic differential evolution algorithm

This paper describes a new method to check for defects in railway tracks for improving passenger safety and comfort. The irregularities in the railway tracks are the fundamental cause of vibration, and different research projects are currently in progress for optimizing the process. External background noises causes the signals sent from the sensors to be distorted. In order to solve this issue, the Railway track Health Monitoring system uses a Dynamic differential Evolution algorithm (RHMDE) for identifying defects in railway tracks. Micro Electro Mechanical System (MEMS) accelerometers are mounted vertically and horizontally on the bogie and axle-box for sensing abnormalities. To locate the irregularities, a new method is included in the proposed RHMDE method. It automatically updates the location of an abnormality even if the signal from the Global Positioning System (GPS) is absent. Four different railway track problems were used for the experimental study, and the time and frequency domain responses were studied. The experimental setup of the proposed RHMDE is tested and compared the Chaos Particle Swarm Optimization (CPSO) and Genetic Algorithm (GA). The experiment results from the experiment prove that the proposed RHMDE method is the superior method for detecting faults in railway tracks. The RHMDE method will greatly improve the quality of railway transportation through detecting the track faults effectively and consistently.

A dynamic differential evolution algorithm for the dynamic single-machine scheduling problem with sequence-dependent setup times

The paper studies the single-machine scheduling problem with sequence-dependent setup times under dynamic environment. In this problem, jobs arrive over time and all the information on a job is unknown in advance. The objective is to find a feasible schedule such that the maximum lateness is minimised. This problem represents a very important production scheduling problem but remains under-represented in the literature. To solve the problem, we propose a dynamic differential evolution algorithm that considers the new environment and the previous environment simultaneously. The performance of the algorithm is enhanced by two heuristics for generating efficient initial solutions, a local search procedure and a speed-up method. To evaluate its performance, the proposed algorithm is tested on 1000 instances from the literature. The computational results demonstrate that the proposed algorithm is highly effective as compared to the methods known from the literature.

Synthesis of linear sparse arrays based on dynamic parameters differential evolution algorithm

Synthesis of antenna arrays is a momentous technology possessing many advantages and broad prospects. Dynamic parameters differential evolution (DPDE) with two modifications is proposed for the synthesis of linear sparse arrays (LSA) which aims at reducing peak side-lobe level. One modification is that the scaling factor decreases monotonically over iterations, making exploration in the later period more efficient and accuracy. The other modification is that the crossover probability changes dynamically during iterative process through a novel self-adaptive strategy. Experiments of checking on benchmark functions and simulations of synthesising LSA validate that DPDE performs better than reference methods convincingly.

Design and optimization of energy-efficient liquid-only side-stream distillation configurations using a stochastic algorithm

The liquid-only side-stream option is sometimes preferred to replace the vapor–liquid interconnection in thermally coupled distillation to achieve similar energy efficiency but easier controlling. This work enumerates all possible liquid-only type distillation sequences for ternary systems before a fair comparison of them, through a case study of equimolar BTX (benzene–toluene–xylene) mixture. The optimization problem can be formulated as a mixed integer nonlinear programming (MINLP). A self-adapting dynamic differential evolution (SADDE) algorithm is used to address its difficulties from inherent nonlinearity and nonconvexity. This stochastic optimization algorithm, written in MATLAB, is linked to Aspen Plus to ensure calculation accuracy. Evaluation and comparison of ten configuration varieties verify the outperformance of liquid-only type configurations over simple distillation sequences with main equipment in common, saving both in energy and economic cost.