Research On Optimal Scheduling Approach Of Large-scale SWRO Desalination Plant

Reverse osmosis desalination technology because of its investment, little energy consumption,convenient operation, easy to control, is recognized as the mainstream of seawater desalinationtechnology in the21st century, after forty years of development, has made rapid progress. Reverseosmosis desalination costs are still higher compared to the municipal ordinary water cost. In reverseosmosis desalination costs, energy costs almost half, how to reduce the production costs becomethe research focus of scholars both at home and abroad. In this paper, on the basis of literaturereview, the establishment of a reverse osmosis desalination system optimization scheduling model,and according to the characteristics of the system optimization model, more in-depth studies thealgorithm of the model.The main contents of this paper include the following aspects:In a typical large multi-machine parallel reverse osmosis desalination project as the researchobject, introduces the main processes, each phase of the unit process and equipment is discussed.Through the analysis of the process of reverse osmosis desalination system, establishes systemoperation cost minimizing the objective function of the scheduling model, established the constraintcondition and boundary conditions, using the penalty function method dealing with the constrainedoptimization problem.Based on optimal scheduling model, genetic algorithm and differential evolution algorithm areused to optimize scheduling. Study the impact on the optimal value of population size and mutationfactor and crossover factor in two algorithms. On the basis of theoretical research, an improveddifferential evolution algorithm is proposed to solve the optimization scheduling model. Theimproved differential evolution algorithm needs further improvement due to the miss that does notconsider the problem of unit input-output ratio in the process of solution, and re-optimization themodel to optimize operation. The problem of input and output was well solved through thedifferential evolution algorithm which was further improved. Using the simulation schedulingresults of genetic algorithm and differential evolution algorithm and its improved algorithm throughanalyzing the optimal performance. Results show that the differential evolution algorithm in theability to search the global optimal solution is superior to genetic algorithm, and the duration of theoperation was obviously shorter than genetic calculation algorithm. For differential evolutionalgorithm in this paper, a hierarchical improved differential evolution algorithm for optimaloperation model, the results verify the effectiveness of the algorithm. Finally, on the basis of theoretical study, an optimization of scheduling software based on theuser demand was designed according to the improved differential evolution algorithm and andcarries on the functional test.