By using more superior algorithm of interchange insertion method the Travel Salesman Problem is solved simply and directly. The idea that is easier to find out the global optimum solution by the way of seeking out the partial optimum solution from solution groups is put forward.
However, signals take effect by the switch of signal phases, which introduces integer control values into the traffic control problem, so the traffic control problem becomes a large-scale mixed integer programming problem, the computation of which requires an exponential-complexity optimization algorithm.
The model employs the Gradient Descent optimization algorithm as parameters learning to refine fuzzy if-then rules, and a system parameter is used to adjust a trade-off between the interpretability and the correct classification percentages, so that the model has good interpretability and learning capability.
With the definition of a new matching dissimilarity measure, an objective function of clustering algorithm is developed by modifying the common cost function, i.e., the trace of the within cluster dispersion matrix, which is suitable for analysis of data sets with mixed attributes. The genetic algorithms (GA) is employed to optimize the developed objective function for overcoming the drawback of the traditional fuzzy k-means (FKM) algorithm, i.e., sensitivity to the initialization.
In connection with the specific features of the application of complex method to the optimum design in engineering structures, a complex method with "layered" iteration taking the iteration number as the control parameter is proposed. Compared with the standard algorithm of complex method, the method here proposed has the advantages of asking for fewer iterations, having higher speed of convergence and being capable of giving an optimum solution of the whole.
An expert system for long-distance truck transport dispatching is presented. In accordance with the experience of dispatchers, the transportation business of the day is divided into subsets. No-load optimum algorithm is then used to arrange the best transportation routes for the subsets. With the local solutions found for the subsets, optimum global solution is obtained. By the system proposed it is possible to improve the mileage utilization of long-distance truck transportation.
This paper introduces and analyses a kind of new heuristic optimization method——simulated annealing algorithm. It is a general random search algorithm based on Monte-Carlo iterative improvement method. The paper gives a sufficient condition for the convergence of simulated annealing algorithm; analyses the main parameters of the algorithm; and discusses its various applications, such as computer design, image processing, and neural net computation, etc.