Improvement Of Ant Colony Algorithm And Its Application In Multipath Measurement System

21st century,praised as the information age,the main characteristics of today's information technology development are digitalization,networking and intelligence.Using computer to realize the intelligence of information processing has becoming an important sign of this information age.As we all know,besides human society,there are also many social insects,such as ants,bees,etc.Although individuals are limited in ability,they exhibit highly in structured social organizations.In 1990 s,M.Dorigo and V.manieaao put forward a kind of heuristic bionic evolutionary algorithm based on population which is caled Ant Colony Optimization(ACO)by simulating the nature of ants' collective behavior.By studying the ants' foraging behavior,they found that the whole ants collaborate with each other through a chemical caled pheromones to form a positive feedback,gathering ants on the multiple paths to the shortest path gradually.In the past two decades of development,the application of ant colony algorithm has widespread to many fields,such as combination optimization,function optimization,system identification,network routing,robot path planning,data mining and large scale integrated circuit design,and have achieved many good results.In this paper,the basic ant colony algorithm is studied and analyzed,and its application on TSP is simulated.Also for some parameters in ant colony algorithm,such as heuristic factor ?,the expected value of the heuristic factor ?,pheromone evaporation coefficient ? and the pheromone amount Q,simulation experiments are carried out to study the reasonable range of each parameters.As a result,more and more scholars have launched the research on the improvement of ant colony algorithm.So far,the improvement of Ant colony algorithm mainly focus on two aspects: one is the improvement of the ant colony algorithm itself,such as pheromone release modification for improvement,in view of the probability of selection ways for improvement,etc.;the other is to integrate with other intelligent optimization algorithms.Firstly,we introduces two kinds of improved ant colony algorithm,then puts forward the improved algorithm,in which local search and weight parameters was added,making the algorithm achieve a balance.In this way,not only can make the search space as large as possible,to find the optimal possible solution;but also,according to the pheromones left by ants,can make the Ant colony algorithm put the focus of search on the path which have higher pheromone concentration at the same time,thus converges to the global optimum with a larger probability.The simulation results show that the performance of the improved algorithm has improved a lot.In this paper,the ant colony algorithm,is applied in the west-east gas terminal control system.West-east gas yard control system is a kind of automatic test system which is used in the gas transmission pipeline system,including circuit testing device,control circuit and circuit switches.The basis of circuit module can replace the switch terminal row.Test module make base module circuit disconnect,then form a new test loop.This method make the channel test mass,simplification,and in the process of the test,it is more accurate and efficient.As for multipath measurement system,in this paper,ant colony algorithm was applied to the system path planning problem of the sensor,I obtained the optimal path graph by doing experiments.In practical engineering,the optimal path as guider for engineers,not only greatly improve the efficiency of planning,but also save a large number of materials.It is very important to guide the actual project.