This article introduces a model-based reinforcement learning (RL) approach for continuous state and action spaces. While most RL methods try to find closed-form policies, the approach taken here employs numerical on-line optimization of control action sequences. First, a general method for reformulating RL problems as optimization tasks is provided. Subsequently, Particle Swarm Optimization (PS...

With the rapid development of swarm intelligence research field, a large number of algorithms in swarm intelligence are proposed one after another. The strong points and the drawbacks of a specific swarm intelligence algorithm becomes clear to be seen when the number of its application increases. To overcome the handicaps, some hybrid methods are invented. In this review, three hybrid swarm int...

This paper proposes a combination of particle swarm optimization (PSO) and Q-value based safe reinforcement learning scheme for neuro-fuzzy systems (NFS). The proposed Q-value based particle swarm optimization (QPSO) fulfills PSO-based NFS with reinforcement learning; that is, it provides PSO-based NFS an alternative to learn optimal control policies under environments where only weak reinforce...

In this paper, an innovative hybrid recursive particle swarm optimization (HRPSO) learning algorithm with normalized fuzzy cmean (NFCM) clustering, particle swarm optimization (PSO) and recursive least-squares (RLS) is proposed to generate radial basis function networks (RBFNs) modeling system with small numbers of descriptive radial basis functions (RBFs) for fast approximating two complex and...

Comprehensive learning particle swarm optimization (CLPSO) is a powerful state-of-the-art single-objective metaheuristic. Extending from CLPSO, this paper proposes multiswarm CLPSO (MSCLPSO) for multiobjective optimization. MSCLPSO involves multiple swarms, with each swarm associated with a separate original objective. Each particle's personal best position is determined just according to the c...

The optimization of nonlinear functions using particle swarm methodology is described. Implementations of two paradigms are discussed and compared, including a recently developed locally oriented paradigm. Benchmark testing of both paradigms is described, and applications, including neural network training and robot task learning, are proposed. Relationships between particle swarm optimization ...

In this paper, a recurrent functional neural fuzzy network (RFNFN) with symbiotic particle swarm optimization (SPSO) is proposed for solving identification and prediction problems. The proposed RFNFN model has feedback connections added in the membership function layer that can solve temporal problems. Moreover, an efficient learning algorithm, called symbiotic particle swarm optimization (SPSO...

This study presents an adaptive two-layer particle swarm optimization algorithm with elitist learning strategy (ATLPSO-ELS), which has better search capability than classical particle swarm optimization. In ATLPSO-ELS, we perform evolution on both the current swarm and the memory swarm, motivated by the tendency of the latter swarm to distribute around the problem’s optima. To achieve better co...

This study proposes a novel bacterial foraging swarm-based intelligent algorithm called the bacterial foraging particle swarm optimization (BFPSO) algorithm to design vector quantization (VQ)-based fuzzy-image compression systems. It improves compressed image quality when processing many image patterns. The BFPSO algorithm is an efficient evolutionary learning algorithm that manages complex glo...