Background
In recent years, global climate is increasingly warmed, which causes the problems of ice melting in the north and south polar sea, sea level rising and the like. Meanwhile, in summer, the conditions of the polar sea area can meet the normal passing of the ship. With the development of ice breaking technology and ship technology, the voyage of China to the east bank of North America through a northwest route and the voyage of China to Europe through a northeast route are greatly shortened, and the transportation cost of ships is saved. The abundant mineral resources in the polar region attract the research of scientific research teams of various countries, and in recent years, more and more ships start to drive to the polar region.
At present, due to the fact that a map open platform comprises a Baidu map, a Google map and a Mapbox, the map open platform is defaulted to be a Mokator projection, polar region projection is lacked, developers are lacked in polar region map layer information, APP is not used for displaying, analyzing and predicting polar region ice information, and sailors cannot clearly know sea ice intensity and depth information around a ship. Due to the lack of a navigation and path planning system suitable for polar region sea areas, the selection of the ship in the navigation course of the polar region ice region is completed to a great extent according to the experience of a navigator, and the navigation safety cannot be effectively guaranteed. The intelligent navigation APP system suitable for polar region ice region navigation is researched and has the first task of collecting sea ice information including sea ice intensity, sea ice thickness and the like, overlapping and fusing ice information data and an electronic chart, and providing the ice information of the surrounding sea region for ships in real time. And planning the target path on the basis, thereby ensuring the normal and efficient passing of the ship. Any reliable tool that supports ice path planning while taking into account physical and operational constraints is very popular in the marine world. Therefore, the intelligent navigation APP system suitable for polar ice region navigation has important significance for the development of follow-up work. In order to develop an intelligent navigation APP system suitable for navigation in polar ice regions, a suitable polar sea ice condition fusion display method is required.
Disclosure of Invention
Due to the fact that the map open platform Baidu map, the Google map and the Mapbox default to the Mokator projection, the polar region projection is lacked, and polar region information cannot be displayed completely. Using ArcGIS to define and project a complete polar region map picture, enabling the picture to correspond to spatial position information, carrying out projection setting by using EPSG (expanded position system) 3571, adding position points to a map layer, enabling a geographic position name to correspond to longitude and latitude, correcting the precision of the longitude and latitude, stopping adding when the error of the longitude and latitude is less than 0.0001 degree, finishing the manufacture of a polar region base map, projecting a vector map address in a server in a polar region, and displaying the polar region projection map layer in a platform in an overlapping mode.
And (2) carrying out normalization processing on the sea ice density and sea ice thickness data files downloaded by the FTP, wherein y is (x-MinValue)/(MaxValue-MinValue), x and y are respectively a value MaxValue before and after conversion, and MinValue are respectively a maximum value and a minimum value of the sample, converting the data into a decimal between (0 and 1), eliminating adverse effects caused by singular sample data, limiting the data in an easy-to-process range, corresponding an output result to an RGB value, and representing the sea ice density and the sea ice thickness by using color saturation. Traversing the matrix, and marking the representative sea ice density and the sea ice depth RGB values corresponding to the longitude and latitude in the polar region projection base map.
Forecasting sea ice condition data by using data mining statistical learning, sorting real-time ice condition and historical ice condition data, corresponding a longitude and latitude matrix to a grid, and establishing a model by using a causal influence network algorithm:
wherein Y represents the sea ice density or depth of a month to be predicted,
representing j climate variables of the ith observation point of K months before the month to be predicted, wherein the climate variables are sea surface temperature, sea surface air pressure and sea surface wind speed,
is composed of
Regression coefficient of Z
-k Representing sea ice concentration or thickness k months before the month to be predicted, q
-k Is Z
-k The regression coefficient of (1), wherein O represents the number of data months, M is the number of grid data, N represents the number of the climate variables, and epsilon is a constant; and comparing the predicted data with the real data, gradually improving the prediction accuracy and providing the universality of sea ice information time.
The method comprises the steps of utilizing a Bayesian network to provide navigation risk reminding for a ship in the process of sailing, obtaining a dependency relationship between nodes by expert investigation and historical experience data of the node as a factor causing an accident of the ship in the process of polar navigation, carrying out correlation analysis on bivariates by adopting a Pearson product moment correlation coefficient method, making the nodes to complete correlation analysis between the nodes, constructing a structure diagram of the Bayesian network in the process of polar navigation risk, and providing navigation risk reminding for the ship in the process of sailing.
Detailed Description
The following further describes embodiments of the present invention with reference to the drawings.
As shown in FIG. 1, the intelligent navigation APP system is suitable for polar region sea ice condition fusion display and data mining and extraction technology, multi-objective optimization algorithm, ship safety analysis technology based on data driving and the like to form navigation in the polar region ice region.
As shown in figure 1, the ice condition display module adopts ArcGIS to define and project a complete polar region map picture to enable the picture to have spatial position information, the projection setting is carried out according to EPSG:3571, the geographical position name corresponds to the longitude and latitude by a method of adding a position point, the addition is stopped when the error of the detected data longitude and latitude is less than 0.0001 degree, and the calibration of the precision of the longitude and latitude is completed. And projecting the vector diagram address in the polar region in the server, and superposing and displaying the polar region projection layer on the platform. And (3) carrying out normalization processing on the sea ice information data, wherein y is (x-MinValue)/(MaxValue-MinValue), x and y are respectively a maximum value and a minimum value of the sample before and after conversion, the data are converted into decimal numbers between (0 and 1), the processed result corresponds to an RGB value, and the saturation of the color represents the sea ice concentration and the sea ice thickness. And marking the corresponding color of the data in the polar projection base map.
As shown in fig. 1, the ice condition prediction module collects and collates ice condition data by using a data mining statistical learning technique, corresponds the longitude and latitude matrix to the ice condition data, and establishes a model by using a causal influence network algorithm:
wherein Y represents the sea ice density or depth of a month to be predicted,
representing j climate variables of the ith observation point of K months before the month to be predicted, wherein the climate variables are sea surface temperature, sea surface air pressure and sea surface wind speed,
is composed of
Regression coefficient of Z
-k Representing sea ice concentration or thickness k months before the month to be predicted, q
-k Is Z
-k The regression coefficient of (1), wherein O represents the number of data months, M is the number of grid data, N represents the number of climate variables, and epsilon is a constant; and comparing the predicted data with the real data, gradually improving the prediction accuracy and providing the ice condition prediction information.
As shown in fig. 1, the route planning module performs route planning by combining meteorological conditions on the basis of sea ice information fusion display, and provides optimal routes under different preferences for users, so that the optimal route for ship navigation no longer depends on the experience of a navigator, and the risk of ship navigation in an ice region is greatly reduced.
As shown in fig. 1, the risk reminding module adopts a ship safety analysis technology, provides a sailing risk reminding for a sailing ship by using a bayesian network, obtains a dependency relationship between nodes through expert survey and historical experience data, performs a correlation analysis on bivariates by using a pearson product moment correlation coefficient method, makes nodes to complete the correlation analysis between the nodes, constructs a structure diagram of the bayesian network for the polar sailing risk, and provides a sailing risk reminding for the sailing ship.
The administrator function module checks the user data, checks and updates the sea ice data on time, and ensures that the system can normally display the sea ice condition and predict the sea ice relatively correctly. And analyzing and updating the ship route data to ensure the reasonability of route planning.
The system interface is simple and has simple function, the optimal route for navigation can be obtained only by inputting destination coordinates or place names, the established data resource library updates the sea ice condition data and the channel information data on time, and the stable operation of the system is guaranteed.
It is to be understood that the specific embodiments of the present invention are illustrative of the principles and applications of the present invention. But not limited to, an embodiment, and other solutions that may be made according to the invention while still being encompassed by the scope of the invention are sought after in the experimental spirit.