基于场景集的危险气体检测报警仪选址方法

传统以典型或等概率泄漏场景为代表开展危险气体检测报警仪选址的做法未能体现装置真实风险,报警仪的探测效率较低。提出一种基于场景集的危险气体检测报警仪选址方法。融合泄漏源工况和风场等重要随机因素生成可能发生的泄漏场景,并定量预测场景发生概率。以概率加权法建立报警仪选址方案的定量评价指标,结合报警仪数量约束构建随机优化模型。结合气体扩散经验模型定义泄漏场景风险指数,在保证风险覆盖率的前提下,优选最大可信场景进行CFD数值模拟。基于量子粒子群算法实现优化模型的定量求解。以某柴油加氢装置为例,采用该方法进行硫化氢检测报警仪选址优化。结果表明,通过该方法获得的最优选址方案与原布置方案相比可较大程度地提高探测效率。

An approach to placement optimization of gas detectors based on leakage scenario set

The traditional gas detector placement method using typical or equal probability leakage scenarios can not reflect the actual risk, and the detection efficiency is also low. Basing on the approximate real leakage scenario set, this article proposes a new approach to gas detector placement optimization. It combines leaking condition and wind field to build leakage scenes that will probably occur and predicts the probability of each leakage scenario quantatively. The probability weighting method was applied to establish the quantitative evaluation index for the detector layout. With the constraint of detector number, the stochastic optimization model was developed. The gas leakage empirical model was also integrated to define the risk index for leakage scenario and the most credible scenario was selected to perform a CFD-based dispersion simulation. To obtain the optimal gas detectors placement efficiently, the quantum particle swarm optimization(QPSO) was employed to solve this model. The method mentioned above was applied to the optimization of hydrogen sulfide detectors for a diesel hydrogenation unit. The results demonstrate that the optimal hydrogen sulfide detectors placement can significantly improve the detection efficiency compared with the original layout.