Soft sensors in fluid catalytic cracking

Through analyzing conventional GA, a new real-coded GA based on annealing chaotic mutation operator is proposed. By introducing the intrinsic stochastic property and ergodicity of chaos movement and variable evolutionary rate, this algorithm can better simulate the process of biologic evolution, and possess the better hill-climbing ability. And it adaptively changes the operating order of evolutionary operators in the different evolutionary stage. So it overcomes the shortcoming of premature convergence and stagnation, and effectively solves the problem of global convergence. Compared with some self-adaptive GA, the test results show that this algorithm is easy to be implemented, and its efficiency is higher in the rate of convergence, accuracy and reliability, so it is effective for optimization problem