基于粒子群优化支持向量机的边坡稳定性预测

摘要

为了提高边坡稳定性预测的精度,保障边坡工程的安全,提出基于粒子群优化算法支持向量机的预测模型.采用粒子群优化算法不断进行搜索迭代获取支持向量机模型的最优参数,避免了支持向量机人为选取参数的盲目性和随意性.通过Matlab编程,应用实例证明:该模型的预测精度较高,预测样本的平均相对误差为3.581 9％,计算速度较快,优于改进的BP算法、GA-BP算法和改进支持向量机算法,在实际的工程应用中有着良好的应用前景.%To improve the precision of the predication of slope stability and ensure the safety of slope work, the prediction model based on the vector machine with particle swarm optimization support, i.e.making use of PSO algorithm to constantly carry out search iteration to gain the optimal parameter of the support vector machine, which avoids the blindness and randomness that the parameter is artificially selected by support vector machine.Via matlab programming, the application case verifies that the prediction model is with relative high precision, the average relative error of the prediction sample is 3.581 9％, it is with relatively fast calculating speed, superior then the optimized BP algorithm, GA-BP algorithm and improved support vector machine algorithm.So that, the model is with good application prospect in the actual engineering application.