Abstract:

Chemical processes are usually high-dimensional multivariable systems. When designing the bottom regulatory PID control system, the traditional subjective analysis methods relying on control engineers are difficult to be applied and fail to the optimal control configuration, so a simple and fast computer optimization algorithm is urgently needed. So, an optimization algorithm for decentralized control loop configuration is proposed based on genetic algorithm. In this algorithm, the structure matrix is encoded in string form and the variable pairings are transformed into an integer array, then through choosing the right genetic operator the operations involving crossover, mutation and selection are applied on the integer array. The optimal variable pairing scheme under unconstrained condition is obtained. Moreover, the genetic operator is improved and adjusted according to the constraint conditions. Based on the optimal pairing scheme under unconstrained conditions, the optimal variable pairing scheme with constraints can be obtained by only using mutation and selection operators with a faster calculation speed. Finally, the validity and correctness of the algorithm are tested by case studies involving TE process.