Page 23 - Read Online
P. 23

Zhu et al. Intell Robot 2022;2(3):200­222  I http://dx.doi.org/10.20517/ir.2022.13  Page 216


               4. CHALLENGES AND PROSPECTS OF UUV MOTION PLANNING AND TRACKING CONTROL
               The motion planning and tracking control of the UUV has a promising future in the maritime projects of
               underwater rescue, detection, investigation, tube pavement, creature study, and military strategy. Hence, there
               is still a large requirement for thorough and systematic research in this field. In this section, the challenges and
               prospects of the UUV motion planning and tracking control for underwater navigation are listed and analyzed.


               4.1. Multi­UUV collaboration
               In this section, possible future studies on the multi-UUV collaboration are given, mainly divided into the
               multi-UUV collaboration considering the environmental effect and heterogeneous vehicles and the hunting
               of dynamic targets.

               4.1.1. Environmental effect and heterogeneous vehicles
               Most researchers discuss the UUV assignment in an ideal underwater environment and regard the vehicle as
               a pure particle, which lacks the consideration of the practical condition of the UUV operation. Therefore, the
               complex environmental factors such as ocean currents effect, unpredictable seamounts, and various moving
               obstacles should be involved in further studies. Moreover, the UUV system contains uncertainty that cannot
               beaddressed initially, suchthat theheterogeneousvehicles ofdifferentmodel parameters, navigating velocities,
               or safe distance are required to be studied. In addition, the topics of the formation control of heterogeneous
               vehicles between UUVs and USVs (or even unmanned aerial vehicles) have become more attractive, as the
               USV can help to instruct the UUV in real-time positioning through its less-affected GPS system and efficient
               communication above the water surface [129,130] .


               4.1.2. Dynamic targets
               Most current UUV motion planning and tracking control studies concentrate on tackling static targets, such as
               theunderwatersearchofcrashedvehicles, yetthefollowingorhuntingofdynamictargetsforUUVsisacrucial
               issue in the military defense for the marine system [131] . The following or hunting of dynamic targets covers the
               topics of dynamic task assignment, intelligence of the moving target, path planning while chasing the target,
               and containment of the target completed by the UUVs, which are still waiting to be further investigated.


               4.2. Efficient underwater positioning and path planning
               The developed underwater positioning methods of real-time efficiency are required, as the conventional po-
               sitioning and navigating systems such as GPS are not valid due to the poor communication affected by the
               underwater environment. In addition, path planning in the underwater condition is the most vital part of
               UUV motion planning, where some innovative intelligent planning methods can be applied in this field to
               improve efficiency, such as reinforcement learning.

               4.3. Robust underwater trajectory tracking
               Owing to the complexity and uncertainty of the underwater environment and the nonlinear UUV system, the
               robustness of UUV trajectory tracking has to be advanced to guarantee the UUV navigates as desired. There-
               fore, the problems that affect the robustness of UUV trajectory tracking such as excessive speed references,
               actuator saturation, and thruster damages are worth investigating [132] .

               4.4. Real­time underwater recognition
               Current approaches that can attain the identity information of the underwater targets are limited due to the
               low invisibility and unpredictable obstacles of the surrounding environment; the inefficient communication
               through the fluids in the deep-sea area also restricts the real-time recognition of the underwater target, thus
               affecting the real-time UUV navigation to unknown targets. Hence, the advanced underwater target detection
               techniquessuchasthemulti-sensorinformationinfusionapproachanddeeplearning-basedimagerecognition
               can be applied in this field to process and achieve the results in time [133–136] .
   18   19   20   21   22   23   24   25   26   27   28