Immune Culture Algorithm And Its Application In The Fault Diagnosis Of Chemical Industry

With the advancement of science and technology, the production systems of chemical industry is developing toward large and complicated direction. Once the equipments go wrong, it will bring great losses to the enterprise and personnel, so we need high-performance fault diagnosis method more and more. In this paper, a new type of intelligent evolutionary algorithm-the culture algorithm and immune clone selection algorithm was studied, a culture algorithm based on the immune clone selection theory was proposed, and the beliefs space of cultural algorithm was improved to update their knowledge and acceptance function. Through nine kinds of function test performance, compared with the immune algorithm and the culture differential evolution algorithm, the results show that the immune culture algorithm had batter performance. And it was been used in parameter optimization for support vector machine, taking into account the curse of dimensionality of high-dimensional data affects the recognition rate of the fault diagnosis, fault feature selection method based on immune cultural algorithm was proposed, the application of these methods go to the TE process simulation model and the actual production process ethylene cracking furnace fault diagnosis, concluded by simulation, fault diagnosis method based on immune cultural algorithm has a very high diagnostic fault diagnosis in the chemical industry, especially the multi-classification accurate rate, reflecting the advantage of immune cultural algorithm.