为提高基于模糊神经网络的锅炉炉膛受热面结渣预测精度，提出了一种基于广义钟型隶属度函数非线性惯性权重递减调整策略的粒子群优化算法，通过适应度测试函数对比实验、结渣预测实验和预测稳定性分析对现有文献中线性惯性权重递减调整策略（LPSO）、指数型非线性惯性权重递减调整策略(IPSO)和基于广义钟型隶属度函数非线性惯性权重递减调整策略(GJPSO)进行对比分析。研究结果表明：本文所改进的粒子群算法可以有效地改善算法的早熟现象、平衡算法的全局和局部搜索能力、提高算法的收敛效果和稳定性。利用改进后的粒子群算法对模糊神经网络中的权值和阈值进行优化，提高了模糊神经网络的炉膛结渣预测性能。

In order to improve the prediction accuracy of the slagging on the heatresistant surface of boiler furnace based on fuzzy neural network,this paper proposed a particle swarm optimization algorithm based on the nonlinear decreasing inertia weight adjustment strategy of generalized bellshaped membership function,then through the fitness function contrast experiment,slagging prediction experiment and predictive stability analysis,analyzed the adjustment strategies of linear decreasing inertia weight(LPSO),index nonlinear decreasing inertia weight(IPSO) and nonlinear decreasing inertia weight based on generalized bellshaped membership function (GJPSO) in existing literature comparatively.The results show that the particle swarm algorithm proposed in this paper can effectively improve the early familiarity of the algorithm,balance the overall and local search capabilities of the algorithm,and enhance the convergence effect and stability of the algorithm.The weight and threshold in the fuzzy neural network is optimized by the improved particle swarm algorithm,and the furnace slagging prediction performance of the fuzzy neural network is improved.

TP183