7 November 2024 in Journal of Bionic Engineering
7 April 2025 in Evolving Systems
28 March 2024 in Research Square
To address the problems of low convergence efficiency and easy to fall into local extremes in the sand cat swarm optimization algorithm for solving the path planning problem of mobile robots. A hybrid multi-strategy based sand cat swarm optimization (HMSCSO) algorithm is proposed to solve the path planning problem of mobile robots. First, the nonlinear adjustment strategy is used to increase the convergence accuracy of the algorithm; Then, the logarithmic weight strategy is introduced into the position update to balance the exploration and exploitation ability of the algorithm; Next, the alternate selection strategy is used to improve the ability of the algorithm to jump out of local extremes; Finally, the position update formula of Lévy flight is introduced into the algorithm to alleviate the situation that the algorithm falls into stagnation. To verify the effectiveness of the proposed HMSCSO algorithm, 23 benchmark test functions and CEC2022 test functions are selected for comparison with other advanced optimizers. In addition, the HMSCSO algorithm is subjected to ablation experiments in three groups of environments with different obstacles. The experimental results show that the HMSCSO algorithm can effectively improve the solution accuracy.
20 August 2026 in Journal of King Saud University Computer and Information Sciences
Distributed coordination of multi-agent systems must remain reliable despite communication delays and the coexistence of cooperative and adversarial interactions induced by resource contention or misbehaving nodes. The central difficulty beyond conventional formation containment is that, once cooperative and adversarial couplings coexist on a sampled network with nonuniform delays, the closed-loop comparison dynamics lose row-stochasticity and individual transition steps become expansive, so neither standard delayed-consensus analysis nor a single Lyapunov or spectral certificate applies; the contribution is a transformation and a row-sum contraction argument that recover a closed-form resilience guarantee in this regime. This paper investigates resilient time-varying formation containment for second-order discrete-time multi-agent systems with cooperative-adversarial interactions modeled by signed graphs and subject to nonuniform delays. A distributed protocol combining each follower’s instantaneous self-state with delayed neighbor and leader states is designed; an explicit model transformation recasts the closed-loop dynamics as an augmented nonnegative comparison system whose row sums can exceed one due to competition. Using sub-stochastic and super-stochastic matrix theory, closed-form design conditions and a quantitative input-to-state stability bound are established, relating the steady-state formation error to the corrected leader- and sign-induced residual and exposing the trade-off among controller gains, sampling period, and cooperation-competition balance. Numerical experiments up to 144 agents confirm graceful error degradation with competitive intensity, bounded behavior in the super-stochastic regime, divergence when the competition margin is violated, and a clear advantage over cooperative-only baselines.