Research On The Shortest Path Model Based On Priority

The priority shortest path is also called the optimal path.It is one of the fundamental research problems and one of the popular research problems in network model optimization.Many research results have been generated in network model optimization research over the years,but many new problems have been found in practical applications such as communication between space satellites,transportation of items,path planning for intelligent robots and intelligent planning of traffic routes.These new problems have posed new challenges and demands for the study of optimal path problems,and many unique solutions have emerged.This paper is based on the solution method of the neural network framework.It mainly studies the priority shortest path problem in static network environment,the priority shortest path problem in time-varying environment,and the priority shortest path problem in multi-constraint time-varying network environment,and analyzes and discusses these three issues.The detailed research content of the three parts are as follows.1.The prioritized shortest path for static networks is modeled and studied.A forward-wave neural network model(FNN)is proposed that can produce accurate results without training and the neurons in the network can be computed in parallel.In this part of the study,the definition of priority is given first,based on which the priority shortest path is defined,and a comparative analysis with the classical algorithm is performed on several standard data sets with randomly generated arc weights to demonstrate the good performance of the FNN.2.The priority shortest path of the priority time-varying network is modeled and studied.Provides a high priority time-varying neural network(PTNN)that transmits information through automatic waves.The neurons in PTNN have a six-layer architecture,and each neuron of PTNN can run and calculate at the same time in principle,and no training is required.Experiments comparing with other algorithms on several standard data-sets with randomly generated arc weights have demonstrated the good performance of PTNN.3.Research on the construction mode of multi-constraint dual time-varying priority and shortest path problem.Provides an automatic wave neural network(AWNN)that requires no training and can be computed automatically.The problem solved in this part is to find the optimal path with the least cost in the network under the condition of multiple constraints,and the priority and cost of arcs in this network are independent time-varying.Comparative experiments with other algorithms on international standard datasets are conducted to demonstrate the good performance of AWNN.