Research On Spiking Neural P System And Its Application In Combinatorial Optimization

Membrane computing is an important branch of natural computing.The computational model of the membrane computing is called the membrane system,which has great parallelism and distributed characteristics.The membrane algorithm is a popular research in the membrane computing,which mainly combines membrane systems and various heuristic algorithms to solve practical problems,and it is a bridge between membrane computing and application.At present,there are many research achievements on the cell-like and tissue-like membrane algorithm,and the research on the neural membrane algorithm is relatively less.This paper mainly studies the membrane algorithm of the spiking neural p systems.The combinatorial optimization problem is that in a given condition;obtain the combination of variables that make the objective function maximal or minimal.In theory,a combinatorial optimization problem can be found the optimal solution by enumeration method,but with the expansion of the scale,there will be the problem of combinatorial explosion.In recent years,the emergence of heuristic optimization algorithm for solving combinatorial optimization provides a new approach and the high parallelism of membrane system can further improve the efficiency of the heuristic algorithm.This paper combines the heuristic algorithm with the membrane system to solve the combinatorial optimization problem.In this paper,we first propose inhomogeneous weighted spiking neural p systems with local homogeneous,and study its computing power.Secondly,we combine membrane system with heuristic algorithm to solve two typical combinatorial optimization problems.In this paper,the main research contents are as follows:Firstly,we put forward the concept of local homogeneity based on the biological mechanism of central nervous system,and design inhomogeneous weighted spiking neural p systems with local homogeneous.Then we use the designed p system simulating the registration machine working in the production mode and receive mode,and proves the universality of this system.Secondly,an optimized inhomogeneous weighted spiking neural p system with local homogeneous(optimized IHWSNP)was proposed.According to the characteristics that SNP can generate binary language and local homogeneity,we design extended spiking neural p system,and then combine it with particle swarm optimization algorithm formed the optimization IHWSNP system to solve the permutation flow shop scheduling problem.We prove the feasibility of the proposed membrane system in solving the permutation flow shop scheduling problem,and compare it with the basic particle swarm optimization algorithm.The experimental results show that the optimized IHWSNP system is more efficient.Thirdly,we further improved the optimized IHWSNP system.The switch controller is introduced to control the excitation of expanded spiking neural p system cluster;it is mainly implemented by the rules of the neuron in the controller.The simulation results show that the system can effectively solve the Rec class problem in the permutation flow shop scheduling.Fourthly,we changed the rules in switch controller of the optimized IHWSNP system,and combine the membrane system with the improved genetic algorithm to solve the traveling salesman problem of 30 cities.The result shows that the optimized IHWSNP system can shorten the time of algorithm convergence.