阅读说明：本技术 一种航线距离测算方法及系统 (Course distance measuring and calculating method and system ) 是由 吴键 于 2021-07-31 设计创作，主要内容包括：本发明提供一种航线距离测算方法及系统,所述方法包括下列步骤：获得AIS数据集；建立交互感知神经网络,将所述AIS数据集以及时间集输入所述交互感知神经网络,获得加速度基于所述加速度以及时间t至时间t+i之间的差值,计算时间t至时间t+i的第一距离值S-(1)；在所述AIS数据集提取瞬时速度参数,基于多个速度参数建立速度集{V-(0),V-(1),…,V-(i)},基于速度集计算时间t至时间t+i的平均速度基于平均速度以及时间t至时间t+i之间的差值,计算时间t至时间t+i的第二距离值S-(2)；基于第一距离值S-(1)以及第二距离值S-(2),计算时间t至时间t+i的航程值S-(t),多个航程值之和构成航线距离。(The invention provides a method and a system for measuring and calculating route distance, wherein the method comprises the following steps: obtaining an AIS dataset; establishing an interactive perception neural network, inputting the AIS data set and the time set into the interactive perception neural network to obtain the acceleration Based on the acceleration And the difference value between the time t and the time t + i, and calculating a first distance value S between the time t and the time t + i 1 (ii) a Extracting instantaneous speed parameters from the AIS data set, and establishing a speed set { V } based on a plurality of speed parameters 0 ,V 1 ,…,V i Calculating the average speed from time t to time t + i based on the speed set Based on average speed And the difference value between the time t and the time t + i, and calculating a second distance value S between the time t and the time t + i 2 (ii) a Based on the first distance value S 1 And a second distance value S 2 Calculating the range value S from time t to time t + i t And the sum of the plurality of range values forms the course distance.)

1. A method for measuring and calculating route distance is characterized by comprising the following steps:

acquiring AIS (automatic identification system) original data, and preprocessing the AIS original data to obtain an AIS data set;

extracting time parameters from the AIS data set, establishing a time set { t, t +15, t +30, …, t + i } based on a plurality of time parameters, wherein t is the departure time, establishing an interactive perception neural network, inputting the AIS data set and the time set into the interactive perception neural network, and obtaining the acceleration between the time t and the time t + i

Based on the accelerationAnd the difference value between the time t and the time t + i, and calculating a first distance value S between the time t and the time t + i₁；

Extracting instantaneous speed parameters from the AIS data set, and establishing a speed set { V } based on a plurality of speed parameters₀，V₁，…，V_iCalculating the average speed from time t to time t + i based on the speed setBased on average speedAnd the difference value between the time t and the time t + i, and calculating a second distance value S between the time t and the time t + i₂；

Based on the first distance value S₁And a second distance value S₂Calculating the range value S from time t to time t + i_tAnd the sum of the plurality of range values forms the course distance.

2. The method for measuring and calculating the distance of the route according to claim 1, wherein AIS raw data is obtained, the AIS raw data comprises a destination, a starting place, longitude, latitude, sampling time and navigational speed, all data including the same destination, the same starting place and the same ship number in the AIS raw data are extracted to form route data, and the route data are interpolated and corrected by adopting a linear interpolation method to obtain an AIS data set.

3. The method as claimed in claim 2, wherein the mutual perceptive neural network comprises convolutional layers as social tensor extractors, fully connected layers as social feature blenders, encoders LSTM for implementing depth feature merging, decoders LSTM for outputting acceleration at Δ t of the ship

4. A method for route distance estimation according to claim 3, characterized in that the first distance value S from time t to time t + i is calculated₁：

In the formula, V_tAt time t, Δ t is the difference between time t and time t + i.

5. The method of claim 1, wherein the velocity V is measured₀Corresponding to time t, said speed V_iThe corresponding time is t + i, so the average velocity from time t to time t + i is calculated by

6. The method as claimed in claim 1, wherein the flight path distance is calculated by using a plurality of flight path values S_tAnd accumulating to obtain the final range value of the air route.

7. The method as claimed in claim 1, further comprising obtaining final range values of any one of the same vessels at different time periods, and taking an average of the final range values as the course distance.

8. An en-route distance estimation system characterized by performing the en-route distance estimation method according to any one of claims 1 to 5, the system comprising:

the data extraction module is used for accessing the automatic ship identification system, acquiring an AIS (automatic identification system) original data set, and extracting route data from the AIS original data set to acquire the AIS data set;

the neural network module is used for constructing a neural network, the neural network comprises a convolutional layer, a full connection layer, an encoder LSTM and a decoder LSTM, the convolutional layer is used as a social tensor extractor, the full connection layer is used as a mixer of social characteristics, the encoder LSTM is used for realizing the combination of depth characteristics, and the decoder LSTM is used for outputting the acceleration of the ship at delta t

The first data collection module is used for extracting time parameters from the AIS data set and constructing a time set;

the second data collection module is used for extracting speed parameters from the AIS data set and constructing a speed set;

a first distance calculation module for calculating a first distance based on the accelerationCalculating a first distance value;

the second distance calculation module is used for calculating a second distance value according to the speed parameter and the time parameter;

and the voyage calculation module is used for calculating the voyage of the navigation according to the first distance value and the second distance value.

Technical Field

The invention relates to the technical field of route calculation, in particular to a route distance measuring and calculating method and system.

Background

In the process of ship navigation, due to the existence of marine environment interference or avoidance operation requirements, the actual course distance of a ship can have deviation through a theoretical course distance calculated by a speed-time equation, so that the course distance is not calculated accurately.

Disclosure of Invention

The invention aims to provide a route distance measuring and calculating method to solve the problems in the background technology.

The invention is realized by the following technical scheme: the invention provides a route distance measuring and calculating method in a first aspect, which comprises the following steps:

acquiring AIS (automatic identification system) original data, and preprocessing the AIS original data to obtain an AIS data set;

extracting time parameters from the AIS data set, establishing a time set { t, t +15, t +30, …, t + i } based on a plurality of time parameters, wherein t is the departure time, establishing an interactive perception neural network, inputting the AIS data set and the time set into the interactive perception neural network, and obtaining the acceleration between the time t and the time t + i

Based on the accelerationAnd the difference value between the time t and the time t + i, and calculating a first distance value S between the time t and the time t + i₁；

Extracting instantaneous speed parameters from the AIS data set, and establishing a speed set { V } based on a plurality of speed parameters₀,V₁,…，V_iCalculating the average speed from time t to time t + i based on the speed setBased on average speedAnd from time t to time t + iThe second distance value S from time t to time t + i is calculated₂；

Based on the first distance value S₁And a second distance value S₂Calculating the range value S from time t to time t + i_tAnd the sum of the plurality of range values forms the course distance.

Optionally, the AIS raw data is obtained, where the AIS raw data includes a destination, a departure place, a longitude, a latitude, a sampling time, and a navigational speed, all data including the same destination, the same departure place, and the same ship number in the AIS raw data are extracted to form route data, and the route data is interpolated and corrected by using a linear interpolation method to obtain an AIS data set.

Optionally, the mutual perception neural network includes a convolutional layer, a fully-connected layer, an encoder LSTM, and a decoder LSTM, where the convolutional layer is used as a social tensor extractor, the fully-connected layer is used as a mixer of social features, the encoder LSTM is used to implement merging of depth features, and the decoder LSTM is used to output an acceleration of the ship at Δ t

Optionally, a first distance value S from time t to time t + i is calculated₁：

In the formula, V_tAt time t, Δ t is the difference between time t and time t + i.

Optionally, the speed V₀Corresponding to time t, said speed V_iThe corresponding time is t + i, so the average velocity from time t to time t + i is calculated by

Optionally, a plurality of range values S_tAnd accumulating to obtain the final range value of the air route.

Optionally, the method further includes obtaining final range values of any one of the same ships in different time periods, and taking an average value of the final range values as the course distance.

A second aspect of the present invention provides a route distance estimation system that performs the route distance estimation method according to the first aspect of the present invention, the system including:

the data extraction module is used for accessing the automatic ship identification system, acquiring an AIS (automatic identification system) original data set, and extracting route data from the AIS original data set to acquire the AIS data set;

the neural network module is used for constructing a neural network, the neural network comprises a convolutional layer, a full connection layer, an encoder LSTM and a decoder LSTM, the convolutional layer is used as a social tensor extractor, the full connection layer is used as a mixer of social characteristics, the encoder LSTM is used for realizing the combination of depth characteristics, and the decoder LSTM is used for outputting the acceleration of the ship at delta t

The first data collection module is used for extracting time parameters from the AIS data set and constructing a time set;

the second data collection module is used for extracting speed parameters from the AIS data set and constructing a speed set;

a first distance calculation module for calculating a first distance based on the accelerationCalculating a first distance value;

the second distance calculation module is used for calculating a second distance value according to the speed parameter and the time parameter;

and the voyage calculation module is used for calculating the voyage of the navigation according to the first distance value and the second distance value.

Compared with the prior art, the invention has the following beneficial effects:

according to the method and the system for measuring and calculating the route distance, provided by the invention, the actual navigation distance of each route can be calculated by analyzing a large amount of AIS and ship-age information of a ship on the same route and calculating the distance sum of all AIS nodes in a course by taking time t as a node, and the data deviation can be continuously corrected by continuous large-scale calculation, so that the actual navigation distance of each route is measured and calculated, and a good data basis is provided for ship-age planning and dynamic reminding of the ship.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only preferred embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive exercise.

FIG. 1 is a flow chart of a route distance measuring method according to the present invention;

FIG. 2 is a diagram of a flight path distance estimation system according to the present invention.

Detailed Description

In order to better understand the technical content of the invention, specific embodiments are provided below, and the invention is further described with reference to the accompanying drawings.

Referring to fig. 1, a first aspect of the present invention provides a route distance measuring method, including the steps of:

s1, acquiring AIS original data, and preprocessing the AIS original data to obtain an AIS data set;

s2, extracting time parameters from the AIS data set, establishing a time set { t, t +15, t +30, …, t + i } based on the time parameters, wherein t is the departure time, establishing an interactive perception neural network, inputting the AIS data set and the time set into the interactive perception neural network, and obtaining the acceleration between the time t and the time t + i

S3, based on the accelerationAnd the difference value between the time t and the time t + i, and calculating a first distance value S between the time t and the time t + i₁；

S4, extracting instantaneous speed parameters from the AIS data set, and establishing a speed set { V } based on a plurality of speed parameters₀,V₁,…,V_iCalculating the average speed from time t to time t + i based on the speed setBased on average speedAnd the difference value between the time t and the time t + i, and calculating a second distance value S between the time t and the time t + i₂；

S5, based on the first distance value S₁And a second distance value S₂Calculating the range value S from time t to time t + i_tAnd the sum of the plurality of range values forms the course distance.

In the embodiment of the application, the AIS raw data is acquired, for example, the AIS information acquired from the AIS system includes the Chinese ship name, the ship type, the position, the navigation speed, the destination, the departure place, the longitude, the latitude and the sampling time of the ship. The obvious error records in the AIS data are generally in the following categories: (1) the marine mobile service identification code (MMSI) length of a ship is not a 9-digit number or unreasonable record; (2) the longitude and latitude of the ship exceed a reasonable range (if the longitude and latitude are negative values); (3) the navigational speed and the course of the ship exceed a reasonable range; (4) the acquisition time of the ship information exceeds a reasonable range. Meanwhile, AIS data is lost due to equipment aging, transmission system faults and the like, the lost data needs to be processed, and the AIS original data is interpolated and corrected by adopting a linear interpolation method;

and then taking the destination and the departure place as keywords, extracting the course information corresponding to the destination and the departure place, wherein the course information comprises a plurality of sampling information points, and each sampling information point comprises information such as instantaneous speed, acceleration, longitude and latitude of the ship at time t, so as to obtain an AIS data set.

In step S2, the mutual perception neural network includes a convolutional layer as a social tensor extractor, a fully connected layer as a mixer of social features, an encoder LSTM for implementing depth feature merging, and a decoder LSTM for outputting an acceleration at Δ t of the ship

Its mutual sensing accelerationThe expression of (a) is:

wherein the content of the first and second substances,for the purpose of the recorded acceleration of the vessel,for the length of the vessel to be recorded,for the recorded accuracy of the vessel to be recorded,for the recorded latitude of the vessel,for the recorded yaw angle it is,is a repulsive interaction force with the environment, andcan be expressed by the following formula:

in addition, for the overfitting problem that may occur in the neural network model, Dropout method is adopted to process the overfitting problem, and the threshold value is set to 0.5.

In step S3, a first distance value S from time t to time t + i is calculated₁：

In the formula, V_tAt time t, Δ t is the difference between time t and time t + i.

In step S4, the speed V₀Corresponding to time t, said speed V_iThe corresponding time is t + i, so the average velocity from time t to time t + i is calculated by

Further, by average speedAnd calculating a second distance value by the time difference value delta t

In the previous step, the flight path is segmented, that is, the flight path is divided into N small segments by taking the sampling time as an interval, and the distance of each small segment is determined by a first distance value S₁A second distance value S₂The calculation is performed in the following way:

the N voyage values S_tAnd accumulating to obtain the final range value of the air route.

Optionally, the method further includes obtaining final range values of any one of the same ships in different time periods, and taking an average value of the final range values as the course distance.

As an example, the final voyage value of the ship 1 from the port a to the port B in morning No. 6 is calculated through the above steps while the final voyage value of the ship 1 from the port a to the port B in morning No. 8 is continuously calculated, and the average value of the plurality of final voyage values is taken as the course distance.

Referring to fig. 2, a second aspect of the present invention provides a flight path distance estimation system that performs the flight path distance estimation method according to the first aspect of the present invention, the system including:

the data extraction module is used for accessing the automatic ship identification system, acquiring an AIS (automatic identification system) original data set, and extracting route data from the AIS original data set to acquire the AIS data set;

the neural network module is used for constructing a neural network, the neural network comprises a convolutional layer, a full connection layer, an encoder LSTM and a decoder LSTM, the convolutional layer is used as a social tensor extractor, the full connection layer is used as a mixer of social characteristics, the encoder LSTM is used for realizing the combination of depth characteristics, and the decoder LSTM is used for outputting the acceleration of the ship at delta t

The first data collection module is used for extracting time parameters from the AIS data set and constructing a time set;

the second data collection module is used for extracting speed parameters from the AIS data set and constructing a speed set;

a first distance calculation module for calculating a first distance based on the accelerationCalculating a first distance value;

the second distance calculation module is used for calculating a second distance value according to the speed parameter and the time parameter;

and the voyage calculation module is used for calculating the voyage of the navigation according to the first distance value and the second distance value.

In summary, according to the method and the system for measuring and calculating the course distance disclosed by the application, the actual sailing distance of each course can be calculated by analyzing a large amount of AIS and ship-term information of a ship on the same course and calculating the distance sum of all AIS nodes in a course by taking time t as a node, and the actual sailing distance of each course can be measured and calculated by continuously calculating a large amount of data deviation, so that a good data basis is provided for planning the ship term and dynamically reminding the ship.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.