A Two-Dimensional Pollution Source Location Method Based on Adaptive Neural Evolutionary Algorithm

Abstract

The invention provides a two-dimensional pollution source location method based on an adaptive neuroevolutionary algorithm; firstly, the area to be tested is delineated; then a plurality of individual control agents are generated by using an adaptive neuroevolutionary algorithm to move in the area to be detected, and each individual is calculated The fitness function of this iteration is used to obtain the optimal individual in this iteration; repeated execution for many times, the optimal individual in the complete iterative experiment is obtained, and this individual is the optimal neural network; finally, the optimal neural network is used to control the appropriate When the agent moves in the actual environment, the agent can avoid obstacles and accurately locate the source of pollution. The beneficial effects of the present invention are: the technical scheme proposed by the present invention allows the intelligent body to automatically learn, so that it can move towards a place with higher pollution concentration and avoid obstacles; at the same time, the present invention also uses the ratio of remaining energy as an evaluation index, so that the intelligent body can Pollution sources can be located using less energy during movement, which can effectively reduce costs, improve positioning accuracy and reduce energy consumption.

Description

technical field [0001] The invention relates to the field of pollution source positioning, in particular to a two-dimensional pollution source positioning method based on an adaptive neural evolution algorithm. Background technique [0002] Rapid and accurate location of pollution sources is of great significance to improving the environment and reducing health hazards. Traditional pollution source location algorithms can be roughly divided into two categories: one is rough location methods, such as centroid location algorithm, contour line location algorithm, Bayesian location algorithm, etc. This type of algorithm does not need to build a diffusion model, but the positioning accuracy is not high; the other is an inversion algorithm based on a diffusion model, such as maximum likelihood estimation positioning algorithm, least squares positioning algorithm, least unbiased estimation positioning algorithm, etc. Such methods require prior knowledge of the diffusion of polluti...

AI Technical Summary

Problems solved by technology

The location algorithm based on the concentration gradient is based on the regular hexagon location method, and the concentration gradient is used as the judgment condition to realize the location of the pollution source. However, the intelligent body will shuttle back and forth in the pollution belt, and the search distance may be much longer than other methods.

Based on the particle swarm algorithm positioning method, the particle swarm algorithm is used to control multiple agents to quickly locate the pollution source, which improves the positioning efficiency, but it is easy to fall into the local optimal solution

A positioning method based on path planning and concentration gradient, while considering the problem of obstacle avoidance, realizes the positioning of pollution sources, but this method needs to rasterize the environment, and the resulting errors will affect the positioning results

[0004] In addition, most methods do not take factors such as detection errors, obstacles, and energy consumption into consideration, and cannot meet the application requirements of complex environments.

Images