基于改进PSO优化LSSVM的传感器补偿研究

Compensation of sensors based on improved PSOLSSVM

针对最小二乘支持向量机在对传感器进行补偿时,正则化参数和核函数参数对补偿精度影响较大的问题,提出一种利用改进的粒子群优化算法优化最小二乘支持向量机模型参数的传感器补偿方法。该方法利用改进的粒子群优化算法优化最小二乘支持向量机模型的正则化参数和核函数参数,避免了人工选择参数的盲目性,提高了最小二乘支持向量机模型的预测精度。仿真实验表明,在传感器的补偿时,该方法比最小二乘支持向量机模型的补偿精度更高。

When the compensation of sensors are carried out based on Least Square Support Vector Machine （LSSVM）, the choice of two parameters, regularization parameter C and kernel function parameter a, has effect on precision of compensation. For solving this problem, a LSSVM control method based on improved Particle Swarm Optimization （PSO） is proposed in this paper, by which improved PSO is used to optimize two parameters of LSSVM . This method can avoid the blindness of manual parameter choice and enhance forecasting precision. The simulation results show that this method has advantages in dynamic respond and dynamic error compared with LSSVM.