基于自适应模拟退火遗传算法的传感器优化配置研究

针对传感器优化配置组合优化问题,提出了一种基于模态置信度准则MAC的优化算法——自适应模拟退火遗传算法。以模态置信度MAC矩阵的最大非对角元的值极小为目标函数,针对满足传感器数量不变的约束条件问题,提出了二重结构编码遗传算法,并将传统的模拟退火算法改良后,作为一个独立的算子置于遗传算法进化过程中;为了避免出现过早收敛的现象,引入了自适应交叉和变异概率。算例结果表明该混合算法对传感器数目与位置同时实现了优化,得到了满足不同精度要求的传感器优化配置方案。

Application of Optimal Sensor Placement Based on the Improved Adaptive Simulated Annealing Genetic Algorithms

This paper presents an improved adaptive simulated annealing genetic algorithms of optimal sensor placement(OSP) based on Modal Assurance Criterion(MAC).Taking the maximum off-diagonal element of the MAC matrix as a target function and the sensor number as a constraint condition,a dualistic coding genetic algorithm is developed which improve the traditional simulation annealing algorithms and use it as an independent operator of GA.Therefore,adaptive crossover and mutation probabilities are adopted to prevent premature convergence.Example results clearly indicate that the method is feasible in optimal sensor placement and number selection,and optimal placement for different accuracy requirement is obtained.