This paper introduces a new technique called species conservation for evolving parallel subpopulations. The technique is based on the concept of dividing the population into several species according to their similarity. Each of these species is built around a dominating individual called the species seed. Species seeds found in the current generation are saved (conserved) by moving them into the next generation. Our technique has proved to be very effective in finding multiple solutions of multimodal optimization problems. We demonstrate this by applying it to a set of test problems, including some problems known to be deceptive to genetic algorithms.

Erratum: A species conserving genetic algorithm for multimodal function optimization (Evolutionary Computation)

Evolutionary deep learning: A survey

Due to the population-based and iterative-based characteristics of evolutionary computation (EC) algorithms, parallel techniques have been widely used to speed up the EC algorithms. However, the parallelism usually performs in the population level where multiple populations (or subpopulations) run in parallel or in the individual level where the individuals are distributed to multiple resources. That is, different populations or different individuals can be executed simultaneously to reduce running time. However, the research into generation-level parallelism for EC algorithms has seldom been reported. In this article, we propose a new paradigm of the parallel EC algorithm by making the first attempt to parallelize the algorithm in the generation level. This idea is inspired by the industrial pipeline technique. Specifically, a kind of EC algorithm called local version particle swarm optimization (PSO) is adopted to implement a pipeline-based parallel PSO (PPPSO, i.e., P3SO). Due to the generation-level parallelism in P3SO, when some particles still perform their evolutionary operations in the current generation, some other particles can simultaneously go to the next generation to carry out the new evolutionary operations, or even go to further next generation(s). The experimental results show that the problem-solving ability of P3SO is not affected while the evolutionary speed has been substantially accelerated in a significant fashion. Therefore, generation-level parallelism is possible in EC algorithms and may have significant potential applications in time-consumption optimization problems.

Generation-Level Parallelism for Evolutionary Computation: A Pipeline-Based Parallel Particle Swarm Optimization

https://link.springer.com/content/pdf/10.1007/s11633-022-1317-4.pdf

Evolutionary Computation for Expensive Optimization: A Survey

Optimization problems become increasingly complicated in the era of big data and Internet of Things, which significantly challenges the effectiveness and efficiency of existing optimization methods. To effectively solve this kind of problems, this paper puts forward a stochastic cognitive dominance leading particle swarm optimization algorithm (SCDLPSO). Specifically, for each particle, two personal cognitive best positions are first randomly selected from those of all particles. Then, only when the cognitive best position of the particle is dominated by at least one of the two selected ones, this particle is updated by cognitively learning from the better personal positions; otherwise, this particle is not updated and directly enters the next generation. With this stochastic cognitive dominance leading mechanism, it is expected that the learning diversity and the learning efficiency of particles in the proposed optimizer could be promoted, and thus the optimizer is expected to explore and exploit the solution space properly. At last, extensive experiments are conducted on a widely acknowledged benchmark problem set with different dimension sizes to evaluate the effectiveness of the proposed SCDLPSO. Experimental results demonstrate that the devised optimizer achieves highly competitive or even much better performance than several state-of-the-art PSO variants.

/pdf/stochastic-cognitive-dominance-leading-particle-swarm-36udyuzz.pdf

Stochastic Cognitive Dominance Leading Particle Swarm Optimization for Multimodal Problems

Complex continuous optimization problems widely exist nowadays due to the fast development of the economy and society. Moreover, the technologies like Internet of things, cloud computing, and big data also make optimization problems with more challenges including Many-dimensions, Many-changes, Many-optima, Many-constraints, and Many-costs. We term these as 5-M challenges that exist in large-scale optimization problems, dynamic optimization problems, multi-modal optimization problems, multi-objective optimization problems, many-objective optimization problems, constrained optimization problems, and expensive optimization problems in practical applications. The evolutionary computation (EC) algorithms are a kind of promising global optimization tools that have not only been widely applied for solving traditional optimization problems, but also have emerged booming research for solving the above-mentioned complex continuous optimization problems in recent years. In order to show how EC algorithms are promising and efficient in dealing with the 5-M complex challenges, this paper presents a comprehensive survey by proposing a novel taxonomy according to the function of the approaches, including reducing problem difficulty, increasing algorithm diversity, accelerating convergence speed, reducing running time, and extending application field. Moreover, some future research directions on using EC algorithms to solve complex continuous optimization problems are proposed and discussed. We believe that such a survey can draw attention, raise discussions, and inspire new ideas of EC research into complex continuous optimization problems and real-world applications.

/pdf/a-survey-on-evolutionary-computation-for-complex-continuous-5ekzy6wim3.pdf

A survey on evolutionary computation for complex continuous optimization

The developments of electric vehicle (EV) technology and mobile internet technology have made the EV-oriented ride-hailing service a trend in smart cities. In the service scenario, a high-quality order allocation approach is in great need to quickly process a series of customer request orders, so as to reduce total customer waiting time and transportation cost. To simulate real-world customer-EV allocation scenarios, in this paper, a dynamic EV dispatch (DEVD) model is established by considering multi-source data association from five sources, including customer, vehicle, charging, station, and service. To solve the proposed multi-source data associated DEVD model, a memory-based ant colony optimization (MACO) approach is developed. MACO maintains a memory archive to store the historically good solutions, which not only can be used to update pheromone to guide the search, but also can be used to help the reactions to environmental changes. In response to dynamic changes, a partial reassignment strategy is also proposed to re-optimize some of the assigned customer-EV pairs in the historically best solution. Moreover, an exchange or replace local search procedure is designed to enhance the performance. The MACO algorithm is applied to a set of dynamic test cases with different customer request and EV sizes. Experimental results show that MACO generally outperforms the first-come-first-served approach and some state-of-the-art ACO-based dynamic optimization algorithms.

https://ieeexplore.ieee.org/ielx7/6979/4358928/09730786.pdf

Memory-Based Ant Colony System Approach for Multi-Source Data Associated Dynamic Electric Vehicle Dispatch Optimization

Cold chain logistics (CCL) scheduling is an emerging research problem in the logistics industry in smart cities, which mainly concerns the distribution of perishable goods. As the quality loss of goods that occurs in the distribution process should be considered, the CCL scheduling problem is very challenging. Moreover, the problem is more challenging when the dynamic characteristics (e.g., the orders are unknown beforehand) of the real scheduling environment are considered. Therefore, this paper focuses on the dynamic CCL (DCCL) scheduling problem by establishing a practical DCCL model. In this model, a working day is divided into multiple time slices so that the dynamic new orders revealed in the working day can be scheduled in time. The objective of the DCCL model is to minimize the total distribution cost in a working day, which includes the transportation cost, the cost of order rejection penalty, and the cost of quality loss of goods. To solve the DCCL model, a buffer-based ant colony system (BACS) approach is proposed. The BACS approach is characterized by a buffering strategy that is carried out at the beginning of the scheduling in every time slice except the last one to temporarily buffer some non-urgent orders, so as to concentrate on scheduling the orders that are preferred to be delivered first. Besides, to further promote the performance of BACS, a periodic learning strategy is designed to avoid local optima. Comparison experiments are conducted on test instances with different problem scales. The results show that BACS is more preferred for solving the DCCL model when compared with the other five state-of-the-art and recent well-performing scheduling approaches.

A Buffer-Based Ant Colony System Approach for Dynamic Cold Chain Logistics Scheduling

A numerical model was established by using the 3D lattice method to investigate the synchronous propagation mechanism of multiple clusters of hydraulic fractures in interlayered tight sandstone reservoirs in the Songliao Basin in China. The multi-fracture synchronous propagation model under different geological factors and fracturing engineering factors was simulated. The results show that the vertical stress difference, interlayer Young’s modulus, and lithologic interface strength are positively correlated with the longitudinal propagation ability of multiple hydraulic fractures. The three clusters of hydraulic fractures can have adequate longitudinal extension capacity and transverse propagation range with 15 m cluster spacing and a 12 m3/min pumping rate. The viscosity of the fracturing fluid is positively correlated with the ability of hydraulic fracture to penetrate the interlayer longitudinally but negatively correlated with the transverse propagation length. It is recommended that high viscosity fracturing fluid is used in the early stage of multi-clustered fracturing in interlayered tight sandstone reservoirs to promote hydraulic fractures to penetrate more interlayers and communicate more pay layers in the longitudinal direction, and low viscosity fracturing fluid in the later stage to make multiple clusters of fractures propagate to the far end where possible and obtain a more ideal SRV.

/pdf/multi-fracture-synchronous-propagation-mechanism-of-multi-3uhxtne9.pdf

Multi-Fracture Synchronous Propagation Mechanism of Multi-Clustered Fracturing in Interlayered Tight Sandstone Reservoir

With the increasing complexity of real-world optimization problems, many challenges appear to evolutionary algorithms (EAs). When solving these time-consuming or high-complexity problems, although EAs can guarantee the high quality of solutions, the intolerable time costs will influence their availabilities drastically. Thus, many attempts have been made to overcome that problem. With the rapid development of the distributed computing paradigm and platforms, such as the Message Passing Interface (MPI) and Open Multi-Processing (OpenMP), distributed computing has become readily available and affordable for realizing more powerful EAs. In order to find out whether these platforms have any particular difficulties or preference, whether one of them would be more suitable for EAs, we analyze the performance of different distributed EAs (DEAs) based on different distributed computing platforms, using differential evolution (DE) as an example. Finally, we find out that both MPI and OpenMP have their own superiorities and they can improve the speedup obviously. However, MPI is more suitable for computationally expensive problems and can achieve higher speedup than OpenMP.

Lin Shi

Papers

A survey on evolutionary computation for complex continuous optimization

Memory-Based Ant Colony System Approach for Multi-Source Data Associated Dynamic Electric Vehicle Dispatch Optimization

A Buffer-Based Ant Colony System Approach for Dynamic Cold Chain Logistics Scheduling

Multi-Fracture Synchronous Propagation Mechanism of Multi-Clustered Fracturing in Interlayered Tight Sandstone Reservoir

Experimental Study of Distributed Differential Evolution Based on Different Platforms