Differential Evolution With Multivariate Gaussian Sampling For Sensor Arrangement
Kuiling Du, Gang Tang · 2023
Sensor arrangement problems with the unknown number of sensors and obstacles in an area to be covered by sensors, which are complicated for common evolutionary algorithms to search for the global optimum. This paper presented a new evolutionary algorithm to solve the problem, which utilized hybrid mutation and a specific crossover strategy. Sampling the best individual of the population using the multivariate Gaussian distribution during the mutation process, and the individuals generated by multivariate Gaussian sampling would skip the crossover to selection. It has been proved that the new algorithm has better solution quality and faster convergent speed than classic differential evolution (DE).