This paper considers the problem of the multiple traveling salesperson problem (MTSP) with multiple depots and closed path. We analyzed the advantages and disadvantages of Partheno-Genetic algorithm (PGA) in solving the MTSP. By integrating the reproduction mechanism in the invasive weed algorithm (IWO), we have solved the defect that the local information of individuals in the PGA population may be missing, and obtained the improved Partheno-Genetic algorithm with reproduction mechanism (RIPGA), which greatly improves the solution performance. By comparing the RIPGA with other GAs and other intelligent algorithms on a large number of traveling salesperson problem (TSP) test functions, we have well verified the solution performance of the RIPGA on the optimal solution, indicating that the algorithm can effectively avoid local convergence. At the same time, the algorithm is less affected by parameters and has good stability.

To improve the search efficiency of the unmanned aerial vehicle (UAV) swarm in disaster areas, the target distribution probability graph in the prior information is introduced, and a drone cluster search trajectory planning method based on probability of containment (POC) is proposed. Firstly, based on the concept of probability of containment in search theory, a task area division method for polygonal and circular areas is constructed, and the corresponding search trajectory is constructed. Then, the influence of factors, including probability of containment, probability of detection, and probability of success on search efficiency, is sorted out, and the objective function of search trajectory optimization is constructed. Subsequently, an adaptive mutation operator is used to improve the differential evolution algorithm, thus constructing a trajectory optimization process based on the improved adaptive differential evolution algorithm. Through simulation verification, the proposed method can achieve a full coverage search of the task area and a rapid search within a limited time, and can prioritize the coverage of areas with a high target existence probability as much as possible to achieve a higher cumulative success probability. Moreover, the time efficiency and accuracy of the solution are high.

To weaken or avoid the impact of dynamic threats such as wind and extreme weather on the real-time path of a UAV swarm, a path-planning method based on improved long short-term memory (LSTM) network prediction parameters was constructed. First, models were constructed for wind, static threats, and dynamic threats during the flight of the drone. Then, it was found that atmospheric parameters are typical time series data with spatial correlation. The LSTM network was optimized and used to process time series parameters to construct a network for predicting atmospheric parameters. The state of the drone was adjusted in real time based on the prediction results to mitigate the impact of wind or avoid the threat of extreme weather. Finally, a path optimization method based on an improved LSTM network was constructed. Through simulation, it can be seen that compared to the path that does not consider atmospheric effects, the optimized path has significantly improved flightability and safety.

To solve the problem of selecting drones for passive positioning within unmanned aerial vehicle (UAV) swarm and optimizing corresponding trajectories. This article constructs a method for determining and optimizing the trajectory of UAVs based on an improved particle swarm optimization (PSO) algorithm. Firstly, the time difference of arrival (TDOA) positioning principle was introduced and corresponding algorithm models were organized. Afterwards, the objective function and constraint conditions for selecting drones and optimizing flight paths were designed. The correlation between the optimal solutions of the continuous time optimization problem is used to determine the UAV for positioning. This paper constructs the UAV determination process based on similarity screening. At the same time, combined with the characteristics of the problem to be optimized, the Particle Swarm Optimization (PSO) is improved from three aspects: updating the initial position of particles, improving the iteration formula and setting the adaptive termination condition. This paper further constructs the track optimization process based on improved particle swarm optimization. Through simulation experiments and algorithm comparison, it can be seen that the method constructed in this article can determine the drone used for positioning in real-time and optimize its spatial position. Compared to the selected drones and mainstream passive positioning methods, the method in this article reduces errors by more than 60% and 45%.

In the presence of the main lobe jamming, the performance of the modern radar system would degrade significantly. In this study, a blind source separation (BSS) method based on an Adaptive Opposite Fireworks Algorithm (AOFWA) is proposed to address the problem of ineffective target echo detection when the radar is subjected to main lobe jamming. First, a signal model for BSS applied in the anti-main lobe jamming is provided, and the number of source signals is estimated. Then, an adaptive opposite learning operator is designed to improve the optimization capability and convergence speed of the basic Fireworks Algorithm (FWA). Finally, pulse compression is performed for separated signals to achieve peak detection. The simulation result shows that the proposed algorithm can effectively extract target echo under suppression jamming and deception jamming, and the separation precision is relatively high, and its convergence speed is superior.

Aimed at the bottleneck of Firework Algorithm’s performance and the problem that it is easy to fall into the local extremum, this paper introduces a chaotic map approach and analyzes its statistical characteristics. New hybrid chaotic systems and a chaotic perturbation operator are designed. Based on different chaotic map approaches, a new Chaotic Firework Algorithm (CFWA) is proposed and simulated by typical test functions. The results show that performance of the Chaotic Firework Algorithm is generally better than the original algorithms, and the chaotic hybrid approaches are superior to single chaotic approaches. In order to further verify the application of the proposed algorithm in practical engineering problems, CFWA is applied to Blind Source Separation (BSS) of radar signals. As a result, it does well in sorting observed signals, and has faster convergence and better sorting performance than the traditional algorithms. In this way, the CFWA extends its engineering application.