Data Rectification And Soft Measurement Of Methanol Synthesis Based On Improved Particle Swarm Optimization Algorithm

Methanol is an important organic chemical raw materials and high-quality fuel. In modern industrial processes, measurement errors are inevitable. Data rectification technology is applied to reduce the impact of measurement error. This paper illustrates the particle swarm optimization (PSO), which can implement data reconciliation. According to the principle of particle swarm optimization, velocity variation strategy is proposed to bring out some improved algorithms. The improved algorithm is used to implement data rectification of methanol synthesis unit.The conversion rate of synthesis of methanol is one of the important factors that must be considered in the methanol production process. Based on the analysis of the methanol synthesis process, the paper states major factors related to the conversion rate of methanol synthesis. the paper uses the modeling method based on improved PSO-BP neural network to build the soft sensor model. Based on the neural network, soft measurement models are built for the process of synthesis and transformation of methanol production unit, and use the actual production data to validate the precision of the model, results show that the model can meet the practical needs.The paper presents the software design of the modules of data communication, Data rectification and soft measurement for the methanol synthesis unit.