基于IA-BP优化算法的物料含水率测量技术

为了对微波谐振腔含水率测量结果进行校正，提出一种基于IA-BP优化算法的进化神经网络模型。模型首先利用IA算法，对解群分布多样性的特性进行全局搜索；同时结合BP算法中基于梯度信息指导权值调整的性能，进行局部搜索，进而避免在最优解或次优解附近震荡，并迅速收敛到最优值。结果表明：该优化算法预测精度高，且收敛速度快，具有寻优的全局性和精确性，提高了测量精度。

IA-BP Optimization Algorithm Based Measuring Technique for Moisture of Materials

In order to correct the measuring results of moisture from microwave resonator, the IA - BP optimization algorithm based evolutionary neural network model is presented. In the model, IA algorithm is first used to globally search for diverse characteristics of distribution of solution groups. Meanwhile, the local research is conducted combining with the feature of weight values adjusted under gradient information guide in BP algorithm. Thus, the oscillations near the optimal solution or suboptimal solution are avoided and the optimal value is quickly convergent. The results show that the optimization algorithm features high predictive accuracy, fast convergence speed, global and accurate optimization, thus the measuring precision is enhanced.