S&T Brief: PRC Scientists Analyze 3D BWBUG Pod Optimization
Scientists affiliated with Northwestern Polytechnical University and the PLA Navy Submarine Academy developed an algorithm methodology to optimize the deployment of a swarm of blended wing body underwater gliders (BWBUGs) for purposes of passive acoustic surveillance.
The authors present an optimization model for the deployment of a BWBUG "cooperative system" that balances sensor coverage with communication-related energy consumption.
-
Onboard energy is consumed during underwater communication for both transmission and reception (100 times more energy is consumed for transmission than reception). BWBUGs would communicate with each other to share position and target detection data.
-
Sensor coverage is modeled based on the theoretical BWBUG detection range, understood as a sphere surrounding the vehicle, which is a function of the detection threshold of the onboard passive sonar system and the signal to noise ratio. Other variables incorporated into the model are signal source level, attenuation, and detection probability at a given range.
-
The model's core variables are the number of communication links and the number of potential deployment coordinates (X,Y,Z) within the 3D space.
The paper compares two different methods of optimization algorithm development, HHO and DDHM, proposing the latter, with elements of HHO integrated, as the preferred method for evaluating a 3D BWBUG deployment.
-
Harris Hawk Optimization (HHO) - A heuristic algorithm based on the cooperative predatory behavior of Harris hawks, which communicate and dynamically adjust their attack as their prey attempts to escape. For purposes of optimizing a BWBUG sensor network, the "prey" represents the optimal sensor positioning based on coverage and energy consumption, while the hawks are the varying positions as they improve and approach optimum.
-
Dynamic Decision-Assisted Heuristic Method (DDHM) - The advantages of DDHM are real-time system assessment and adaptation, and the application of decision rules that drive heuristic strategies based on the changing state of the optimization process. It also ensures a more comprehensive search for solutions, and minimizes premature convergence on a suboptimal solution. For this study, the DDHM algorithm is enhanced with opposition-Based Learning (OBL) to further expand the search space and prevent early convergence.
The authors then expand their DDHM algorithm into a multi-objective optimization variant in order to achieve two simultaneous objectives related to BWBUG positioning and communication links:
-
Maximize sensor coverage area within the three dimensional surveillance space, which is expressed as a coverage ratio (CR).
-
Minimize the dispersion of energy consumption data within the 3D BWBUG network topology in order to achieve an optimal balance of energy consumption, which is expressed as a variation coefficient (VC).
The authors then present three cases using DDHM to analyze a BWBUG deployment, ultimately concluding that DDHM is more effective at optimizing for sensor coverage and energy consumption than both HHO and MDDHM.
ANALYSIS:
PRC scientists continue to demonstrate interest in BWBUGs as a viable platform for, among other things, deep ocean passive acoustic surveillance for anti-submarine warfare. However, while the authors' underscore the BWBUGs inherent advantages over conventional UUVs, such as high endurance and low radiated noise signature, there are no BWBUG-specific performance characteristics integrated into the model, suggesting that it is vehicle agnostic - i.e. could apply to any underwater vehicle capable of passive acoustic sensing and underwater communication.
CITATION:
Liang, Q., Huang, H., Huang, B., Hu, S., & Yang, C. (2024). Three-dimensional deployment strategy for a multi-BWBUG cooperative system at deep depths. Ships and Offshore Structures, 1–19. https://doi.org/10.1080/17445302.2024.2391810
