Research on underwater unmanned platform route planning based on marine environment constraints

Since the development of unmanned systems, autonomous underwater unmanned platforms have become an important facility for countries to explore the ocean field which play an important role in both military and civilian. In this paper, the route planning of underwater unmanned platform is carried out from three constraints: the risk that affects the operational effectiveness of the unmanned underwater platform, the navigation resistance and the navigation path length. The simulation results show that the model can deal with three constraints reasonably and plan a suitable route.


Introduction
With vast territory and abundant resources, the marine field is one of the main battlefields between modern great powers.The South China Sea is located in a special place.It is an important security barrier for China and one of the areas of great concern to the international community.Developing and utilizing marine natural resources and safeguarding maritime rights and interests in the South China Sea have become important tasks in the South China Sea.In the face of more and more complex international situation and more and more difficult guard tasks, manned submarines have gradually revealed great limitations.With the development and promotion of unmanned systems, more and more countries attach importance to underwater unmanned platforms [1].Its strong mobility and concealment play an important role in the exploration of resources in the civilian field and the surprise combat in the military field.Therefore, underwater unmanned systems have gradually become an important platform for exploration and combat in the South China Sea.In order to ensure that the underwater unmanned platform reaches the destination smoothly and exerts its effectiveness, the path planning task should be done first [2].In the planning process, we should not only consider the safety and concealment of the underwater unmanned platform, but also consider its energy consumption.Due to the high resistance in water, the energy consumption must be reduced by reducing navigational resistance and path length in the course of route planning.
Global path planning of underwater unmanned platforms in seawater is not much.Xu et al. [3] uses the improved genetic and particle swarm optimization algorithm to plan the route of the underwater robot under the flow field generated by the flow function.Gao et al. [4] uses the improved CD* algorithm to plan the route constrained by global energy consumption, path length, dynamic current environment, and local current velocity.Wang [5] qualitatively describes the influence of different environmental factors on the navigation of underwater vehicles and adopts A* algorithm to evade typical internal wave cline.Most of the above studies are based on single-objective programming, only considering the impact of one element alone.And the impact research of related marine elements is basically qualitative.
In this paper, there are three constraints on the track planning of underwater unmanned platforms.First, the impact of marine environmental factors on the effectiveness of underwater unmanned platforms, including the impact of combat effectiveness, detection effectiveness, safety impact and concealment impact.The four effects are included in the unified quantitative system and considered as the risk degree of the performance of the underwater unmanned platform and as the background of track planning.The second is to consider the navigation resistance of the unmanned underwater platform, in order to reduce the energy consumption during the navigation of the unmanned underwater platform as much as possible and enhance the endurance.The third is to consider the route length of the starting point and the destination point, so as to reduce the planned length as much as possible and improve safety and economy.The method of multi-objective programming is used to balance the three constraints.Route planning can also focus on a certain constraint according to different tasks.

Data
In order to study the effect of marine environment on the effectiveness of underwater unmanned platforms more comprehensively, the common marine environmental factors in the 100-meter depth area of the southern South China Sea are used in this paper.The parameters of the underwater unmanned platform refer to common military UUV in the United States.
In the historical data used, the three-dimensional grid data of temperature, salinity and current data and the two-dimensional grid data of transparency were obtained from the Copernicus Marine Environment Monitoring Service Network (CMEMS) and the spatial resolution of the grid was 1/12°×1/12°.The depth Data for the South China Sea is derived from the NGDC's ETOPO1 dataset, published by the U.S. National Geophysical Data Center, which contains ocean floor topography data with a spatial resolution of 1 '×1'.In addition to the basic data, the density, cline, ocean front and other data can be calculated, and the internal wave comes from the internal wave probability obtained by the team.The date is from February.
Due to the different sources of data, the spatial resolution of all data is consistent through interpolation and the time resolution is the monthly average data.

Method
The planning algorithm adopted in this paper is the eight-direction multi-objective Dijkstra pathfinding algorithm based on raster map [6].First to establish a grid map for environment information, so as to establish environmental information.The essence of this method is to divide the established environment information into units and use a block of cells with the same size to represent it.Each cell is set to impassable or passable, and each cell corresponds to a different risk magnitude and current direction.The moving body moves in eight directions in the movement of the grid map, that is, in principle, each cell point can move to the surrounding eight directions of the cell.But these eight directions don't travel the same distance.Dijkstra's pathfinding algorithm is a path planning algorithm that can use weights.It mainly solves the single source shortest path problem of weighted graph.Its main feature is to give a directed non-negative weight graph and the starting point and the target point specified in the project.Starting from the starting point, the greedy algorithm is adopted to traverse the adjacency node of the closest and unvisited vertex each time until the end point is extended.

Model of the effect of marine environment on the effectiveness of underwater unmanned platforms
Taking the operational unmanned underwater platform as an example, this paper explores the risk of marine environment for the unmanned underwater platform to perform tasks from four aspects: the ability of weapon play, the ability of regional detection, the security and the concealment [7].
Generally, the weapons carried by the underwater unmanned platform are mainly concentrated in the light torpedo class.The faster the torpedo, the less likely it is to be intercepted by the enemy and the more effective it is.In addition to its own driving force, the torpedo speed has the greatest impact on the resistance caused by seawater, and the formula is as follows [8] : where   is navigational resistance.  is the density of sea water.  is the drag coefficient. is the cross-sectional area of the light torpedo.  is the speed of the light torpedo relative to the water.  is related to sea water temperature and flow rate, so the density, temperature and flow rate of sea water are selected as influencing factors.
The detection of underwater unmanned platform is mainly based on sonar system, which is affected by the propagation law of ocean noise and sound line.According to the formula of sound velocity in seawater [9], the sound lines in shallow sea are mainly affected by the temperature gradient.Therefore, temperature gradient and noise are selected as influencing factors.
Large-scale current, ocean front and internal wave in the Marine environment are taken as the main factors to explore the safety impact of underwater unmanned platforms.Due to the small size of the underwater unmanned platform, strong currents can cause damage to its surface and even internal structure.The sudden rise or fall of underwater unmanned platforms caused by sudden change of density of ocean density front seriously threatens its navigation safety.Internal wave is a major hidden danger to the underwater unmanned platform.The smaller internal wave is the uneven pressure on its surface, causing damage to its structure and the larger internal wave can lift it out of the water or drag it to the bottom of the sea.
The concealment of underwater unmanned platforms is explored from two aspects: on the one hand, the influence of sonar is contrary to the above detection capability; on the other hand, the analysis is made from the transparency of sea water.
This paper classifies risk as five levels.Through relevant formulas or expert scores, different marine elements are graded and quantified.Since different marine elements have different impact weights, improved Bayesian networks [10] are used to incorporate different indicators into a unified system to obtain total risk assessment.The basic formula is: (2) where   is the weight of each marine element.v i is the Marine factor impact level.c is the operational effectiveness risk level of the underwater unmanned platform.

Navigational resistance model
According to fluid dynamics, the resistance formula of underwater unmanned platform moving in seawater is as follows [11]: where   is navigational resistance. is the density of sea water.  is the drag coefficient. is the cross-sectional area of the underwater unmanned platform when it runs along the current direction.  is the speed of the underwater unmanned platform relative to the sea water.The energy consumed by the movement is [12]: where  is the energy conversion coefficient.Sea water density, temperature and flow rate are taken into account above.The remaining main factor lies in the Angle between the direction of motion of the underwater unmanned platform and the direction of water flow.The smaller the Angle between the two, the less energy the motion consumes.Therefore, the direction of ocean current should be considered in route planning.

Route planning length model
We think the earth is round.For a given points   (lon  , lat  ),  +1 (lon +1 , lat +1 ), according to the following formula between two points of navigation from [13].

Route planning algorithm model
In general, the geographical map of the study area is converted into a grid.Set the starting point as a yellow cell, the target point as a pink cell, the obstacle as a black cell and the path as a green cell.The general path is obtained as shown in the figure 1: Figure 1 Grid model In this paper, the 100-meter depth area of the southern South China Sea is transformed into uniformly sized blank grids according to the spatial resolution of the data.The grid is set to black in areas less than 100 meters deep, which is absolutely impossible for underwater unmanned platforms to navigate.Each of the other squares contains three constraints: the risk to the effectiveness of the subsea unmanned platform, the Angle between the direction of motion and the current and the total path length.The latter two types of information are judged based on actual movement.Three kinds of constraint information are given different weights.So the cost function of each grid point relative to the previous point is obtained.After determining the starting point and the target point, the eight surrounding directional grids are searched and judged respectively from the starting point.Select the least costly grid to advance and color it green.Then take the current grid point as the center, search the eight points around it to judge the forward point and repeat this process until the search stops at the target point.The result is a series of green grids, which are coupled to the geographic map to get the route we need.

Actual result verification
We set the start and end of the underwater unmanned platform sailing at a depth of 100 meters (from point A to point B).As shown in Figure 2: Or the optimal route is obtained according to the multi-objective constraints, as shown in the Figure 3:     7, the Angle between the motion direction of the underwater unmanned platform and the direction of the ocean current is basically less than 90 degrees, accounting for 85.37% of the entire route, which basically ensures the small energy consumption.
Through calculation, the linear distance of two points A and B is 232.56km, and the total distance of the planned route is 378.3601km.The total distance only increased by 62.69%, which is basically within the acceptable range.The navigation path planned according to the model is compared with the two straight routes A and B: on the one hand, the navigation risk of underwater unmanned platform is reduced and the efficiency is enhanced; on the other hand, the Angle between the movement direction of underwater unmanned platform and the direction of ocean current is reduced and the navigation resistance is reduced.Although the total distance travelled has increased, it is within acceptable limits.In general, the planning algorithm basically meets the navigation requirements of underwater unmanned platforms.

Conclusion
In this paper, the eight-direction multi-objective Dijkstra pathfinding algorithm based on raster map is used to plan the trajectory of underwater unmanned platforms under different constraints and the results are basically in line with expectations.However, the constraint conditions are simplified and the efficiency of the algorithm is too low.There may be so many results that the optimal solution cannot be accurately determined.Subsequently, the constraints will be modified to make them more in line with the complex reality.In addition, the next step should be to optimize the pathfinding algorithm and calculate the optimal result.

Figure. 2
Figure. 2 A straight path from A to B of the underwater unmanned platform According to different tasks, different weights are assigned to the performance risk of the underwater unmanned platform, the Angle between the direction of movement and the ocean current, and the path length.Or the optimal route is obtained according to the multi-objective constraints, as shown in the Figure3:

Figure. 3
Figure. 3 Route planning from point A to point B of the underwater unmanned platformAs shown in the figure, the planned navigation trajectory of the underwater unmanned platform is basically in the low-risk or lower-risk area, and only a few points are located in the marginal area of mediumhigh risk.Compared with the linear navigation shown in FIG.2, in the whole re-planned route, the underwater unmanned platform is in a better Marine environment, which provides a good guarantee for its combat effectiveness.The data of sea current velocity, ocean front, internal wave jump and other factors which have great influence on the safety of the whole route are obtained.As shown in Figure4,5,6.

Figure. 4
Figure. 4 Distribution of ocean current velocity at sailing track points

Figure. 5 Figure. 6
Figure. 5 Distribution of ocean front values at sailing track points

Figure. 7
Figure.7 Comparison between current direction and track directionAs shown in the figure7, the Angle between the motion direction of the underwater unmanned platform and the direction of the ocean current is basically less than 90 degrees, accounting for 85.37% of the entire route, which basically ensures the small energy consumption.Through calculation, the linear distance of two points A and B is 232.56km, and the total distance of the planned route is 378.3601km.The total distance only increased by 62.69%, which is basically within the acceptable range.