Large Field Leveling Earthwork Optimization Based On Evolutionary Algorithm

Earthwork allocation is based on the goal of the minimum allocation costs, with the full consideration to the actual situation at the construction site, to determine the allocation and the transport direction of the amount of earth- rock excavation. When engaging in large-scale civil construction, the cost of earthwork allocation plays a decisive role in the level of the construction benefits, affects the cost, schedule, quality, and the surrounding ecological environment of the entire civil construction projects.In this paper based on the evolutionary algorithms, combined with the principles and characteristics of the earth deployment, according to the mathematical model Earthwork deployment, proposes and prepares a real value code optimization techniques adaptation to the large earthwork allocation program which applied to practical engineering,and proves the feasibility of the optimization techniques.The main contents are as follows:(1) According to features of excavation area, fill area, debit area, spoil area in earthwork construction area, do the debit area as part of the excavation area, spoil area as part of the fill area. According to the balance of excavation area and fill area to establish field level earthwork allocation mathematical model, and refer to the evolutionary algorithms written of mathematical.(2) Simulated the phenomenon of natural biological choice which based on the roulette selection method, and compared the cumulative probability calculated as the fitness with the probability given by roulette to do the choice of survival of the fittest.(3) Taking into account too high cross-rate will destroy the superior genetic population, too low crossover rate will reduce the diversity of the population; too high mutation rate will cause random search algorithm, too low mutation rate will reduce the diversity of the population, so This paper selects dynamic cross mutation operators to simulate randomness, irregularity and dynamics of biosphere. After several debug application of the algorithm, ultimately, the adjustment coefficient of the cross and variation are 0.6 and 0.07.(4) According to the characteristics of the mathematical model, using linear recombination, make the population newly created to meet the requirements of the balance of excavation area and fill area. To do the mutation operation with The population newly generated, arrange the individual chromosomes which will mutate by row and get new populations which meet the requirements of the problem constraints. Using this method to do the crossover and mutation operations, eliminating the need to repair the population, is the innovation of the evolutionary to the algorithm operation, and improve the computational efficiency of the algorithm.(5) To avoid losing the parent outstanding individual genes, introducing the best saving tactics in operation, make the best individual of parent compare and replace the worst individuals of offspring. The individual new got compared with the parent individual, selecting the policy of choosing the best offspring policy, to get new population by selecting the individual of larger fitness.(6) Comparison the written algorithm with linear programming method, prove the effectiveness of the algorithm by way of example, and apply the algorithm to the engineering.