论文部分内容阅读
差分进化算法(differential evolution algorithm,DE)是近年来比较流行的进化算法,被广泛应用于复杂的优化问题中。但常规的差分进化算法存在停滞现象,容易使算法收敛停止,鉴于此本文提出一种改进的差分进化算法,防止了停滞现象也加快了收敛速率并使算法能收敛到全局最优。基于小生境差分进化算法,提出了多维传感器动态解耦方法。该方法克服了原经典方法中对传感器模型精确已知的要求,具有一定的鲁棒性。仿真结果证明了该方法的正确性和有效性。
Differential evolution algorithm (DE) is a popular evolutionary algorithm in recent years and is widely used in complex optimization problems. However, the conventional differential evolution algorithm has the problem of stagnation, which makes it easy to stop the algorithm convergence. In view of this, an improved differential evolution algorithm is proposed in this paper to prevent the phenomenon of stagnation and speed up the convergence rate and make the algorithm converge to the global optimum. Based on niche difference evolutionary algorithm, a dynamic decoupling method for multi-dimensional sensors is proposed. This method overcomes the exact known requirements of the original classical sensor model and has some robustness. Simulation results show that the method is correct and effective.