Research And Application Of Wolf Colony Algorithm

Wolf Colony Algorithm proposed in 2011 is one of the swarm intelligence optimization algorithms, which has been widely used in many fields such as artificial neural network, medical science, optimization triaxial sensor placement, optimization operation of hydropower station and so on. At present, WCA has been obtained good application results in above fields. However, the same as other swarm intelligence optimization algorithms, during the study period, WCA has some problems need to be discussed such as how to improve the performance of the algorithm in optimization problems and expanding algorithm application fields, etc.For expand WCA algorithm theoretical system and practical applications, according to the insufficiency of WCA, it is deeply investigated from theory and application aspects in this paper. In theory, according to a series of optimization problems characteristic, mainly have two aspects of optimization. On the one hand, the structure and key steps of the algorithm are improved to improve its optimal performance in every optimization problem. On the other hand, introduce other techniques made the WCA has handle ability of multimodal function optimization and multi-objective optimization problems. In application, the improved WCA algorithm is applied successfully to solve the Problem of Unmanned Aerial Vehicle trajectory Planning. Concrete contents are as follows.Firstly, for improvement the WCA problems of slow convergence, low solution precision and being easy to fall in local extremum in single objective optimization problem, etc, the internal operating mechanism of WCA are deeply investigated. An improved wolf colony algorithm called modified wolf colony algorithm(MWCA) was proposed to improve the optimization performance. Concrete improvement measurement in the MWCA including: 1、For enhance the explored ability of wolves, the interactive strategy was leaded into scouting behaviors and summoning behaviors. 2、An adaptive beleaguered strategy was proposed for beleaguering behaviors, which made the algorithm has an adjustment function. It can accelerate the algorithm’s convergence rate and convergence precision. Experimental results show that the MWCA algorithm obtains higher solving accuracy, faster convergence speed.Secondly, in order to improve WCA algorithms ability of solving multimodal optimization problems, a niche wolf colony algorithm(NWCA) is proposed combining WCA and the niche technology based on a lot of experiments. On the one hand, the modified niche model can increase population diversity and enhance the capacity of identifying every peak. On the other hand, according to the characteristic of multimodal optimization problems, concrete improvement measurement in NWCA algorithm including: 1、considering the fitness value method of choosing the scouting wolf is not suiting multimodal optimization problems, a new fitness shared value method of choosing the scouting wolf is proposed to increase the scouting wolf population diversity; 2、Using the crowding strategy, confirmed the algorithm evolution direction; 3 、 For renewable mechanism, using an adaptive niche technology to make sure the iterative Population and strengthen the searching ability of in each peak value; 4、For enhance the explored ability of wolves, the interactive strategy was leaded into scouting behaviors and summoning behaviors. An adaptive beleaguered strategy was proposed for beleaguering behaviors. It can accelerate the algorithm’s convergence rate and convergence precision. Experimental results show that the NWCA algorithm can identify each peak accurately and obtain higher solving accuracy.Thirdly, in order to improve the convergence and distribution of WCA algorithms in multi-objective optimization problems, a multi-objective optimization based on Wolf Colony Algorithm(MO-WCA) is proposed. Concrete improvement measurement in the MWCA including: 1、For accelerate the population evolution and avoid the algorithm into local extremum, according to the elite search strategy and the sine function method to improve the scouting behavior, enhance the search ability; 2、Based on the characteristics of multi-objective optimization, combined with the Pareto dominance definition, is proposed based on Pareto dominance relationship self-adapting summoning behavior, effectively promote the evolutionary algorithm; 3、Through the individual fitness adjusting formula, combined with the boundary of individual wolves, is proposed a self-adapting beleaguering behavior to avoid blind search action; 4、Combined with the external population, provided the evolution direction of the population and Retention of high quality individuals. Experimental results show that the MO-WCA algorithm has good convergence and uniformity.Fourthly, aiming at the problem of UAV trajectory planning optimization, a method of equivalent terrain simulation was used to make enemy threatens and terrain obstacles equal to peaks, based on this, the model of trajectory planning of UAV is established. Using modified wolf colony algorithm was adopted to plan the UAV trajectory planning of UAV, whose start point and end point had been known. The experiments show that the MWCA algorithm can effectively shorten the track distance and effectively solved the UAV trajectory planning problem.In conclusion, this topic added the interactive strategy and adaptive strategy based on the WCA algorithm was proposed an MWCA to improve the optimization performance. Combination with the niche technology and crowding strategy, at the same time optimization WCA algorithm search strategy was proposed an NWCA algorithm that was very good to solve the problem of multi peak optimization. Make up the WCA’s blank in the multi peak optimization problems. Combination with the Pareto dominance definition and harmonic crowding distance was proposed an MO-WCA algorithm that was performance good convergence and uniformity in multi-objective optimization problems. Make up the WCA’s blank in the multi-objective optimization problems. For the the problem of UAV trajectory planning optimization, using the MWCA algorithm can effectively shorten the track distance and effectively solved the UAV trajectory planning problem.