基于改进遗传爬山算法的篦冷机熟料换热二次风温故障诊断

Fault Diagnosis of Secondary Air Temperature of Grate Cooler Cement Clinker Heat Transfer Based on Improved Genetic Hill Climbing Algorithm

提出一种利用改进遗传算子优化爬山的算法,能自动调整交叉概率和变异概率,自动选择交叉方式和变异方式,利用改进的交叉算子和变异算子替代爬山算法中的减边算子来扩展其全局寻优能力,利用最大支撑树限制搜索空间来提高搜索效率。仿真实验证明,与爬山算法等算法相比,该算法得到的模型更加准确、最佳得分更高且耗时较短。结合水泥熟料换热过程中的篦冷机现场采集数据,建立了篦冷机工艺参数的故障诊断模型,对二次风温实现了较为准确的故障诊断,具有一定实际意义。

An improved combined algorithm （improved genetic hill climbing, IGHC ） which utilizes the improved genetic operators to optimize Hill-Climbing is proposed. The algorithm adjusts the probability of crossover and mutation self adaptive, selects the way of crossover and mutation automatically, utilizes the crossover and mutation operators replace the edge deletion operator of HC to extend its ability of global optimization, and utilizes the Most Weight Supported Tree restrict the search space to improve search efficiency. Compared with HC and GA（genetic algorithm） , the simulation results have shown that the IGHC algorithm can obtain a model more accurate and rapid, the best score of model is higher than GA and HC. Combined with cement grate cooler operating data in the process of cement clinker heat transfer, the fault diagnosis model of technological parameter can be constructed and a precise fault diagnosis of secondary air temperature in the cement grate cooler can be realized, so this algorithm has a certain practical significance.