An adaptive neighborhood-based search enhanced artificial ecosystem optimizer for UCAV path planning

Abstract

Path planning is an essential component of the unmanned combat aerial vehicle (UCAV) system and is the precondition for achieving tasks such as aerial reconnaissance, monitoring, and fire attack. Its objective is to find a satisfactory route with full consideration of the threat area and constraints, which is a multi-constraint global optimization problem. The work proposes an adaptive neighborhood-based search enhanced artificial ecosystem optimizer (NSEAEO) to address the UCAV path planning problem. The new algorithm uses distance-fitness-based information in the consumption phase to construct an adaptive neighborhood for consumers, who choose the better individuals within the neighborhood for predation. The strategy not only facilitates a thorough search of the problem space but also enhances the global exploration capability of the algorithm. In addition, due to the lack of diversity of the population in the decomposition stage of the AEO algorithm, it is easy to fall into the local extremum. A novel updating decomposition mechanism is designed to dynamically choose the decomposition method according to the threshold in each iteration, which effectively averts insufficient diversity and increases the ability to escape from local extrema. Furthermore, to further improve the search capability of the algorithm, a quadratic interpolation (QI) operator is embedded in the AEO algorithm. The core idea is to choose three agents to match a quadratic function close to the goal function and adopt the extremum of the quadratic function to produce new agents. The above approach achieves a good balance between exploitation and exploration. The experimental results on a series of benchmark functions and UCAV path planning in complex environments demonstrate that NSEAEO outperforms other algorithms in terms of solution quality.