CN117688288A - Sea area ship density non-parameter estimation method - Google Patents
Sea area ship density non-parameter estimation method Download PDFInfo
- Publication number
- CN117688288A CN117688288A CN202311282157.3A CN202311282157A CN117688288A CN 117688288 A CN117688288 A CN 117688288A CN 202311282157 A CN202311282157 A CN 202311282157A CN 117688288 A CN117688288 A CN 117688288A
- Authority
- CN
- China
- Prior art keywords
- ship
- sea area
- ships
- function
- sea
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 19
- 238000005315 distribution function Methods 0.000 claims abstract description 23
- 230000001186 cumulative effect Effects 0.000 claims abstract description 11
- 238000012544 monitoring process Methods 0.000 claims abstract description 10
- 238000012163 sequencing technique Methods 0.000 claims description 3
- ATJFFYVFTNAWJD-UHFFFAOYSA-N Tin Chemical compound [Sn] ATJFFYVFTNAWJD-UHFFFAOYSA-N 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- YOBAEOGBNPPUQV-UHFFFAOYSA-N iron;trihydrate Chemical compound O.O.O.[Fe].[Fe] YOBAEOGBNPPUQV-UHFFFAOYSA-N 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
Abstract
The invention discloses a sea area ship density non-parameter estimation method, which comprises the steps of acquiring corresponding ship numbers of different sizes and different types of ships in a specific sea area as samples through AIS information; constructing a non-parameter estimation model taking a Gaussian function as a kernel function; constructing a sample distribution function approximate cumulative probability distribution function: randomly generating the number p of [0,1], and inputting the number p into a sample distribution function to obtain an integer serving as an estimated value of the number of ships; and updating the model by the supplementary data set, and obtaining the number of the ships in each category at the current moment according to the model to realize the real-time monitoring task of the ships in each sea area. And estimating the number of the ships by adopting a kernel function non-parameter estimation method and empirical distribution approximate overall distribution to obtain a specific sea area ship distribution rule, and providing theoretical support for space-based remote sensing sea area monitoring and channel planning tasks.
Description
Technical Field
The invention relates to the field of space-based remote sensing sea area monitoring and channel planning, in particular to a sea area ship density non-parameter estimation method.
Background
With the development of maritime industry, maritime departments need to control the distribution of the ship in the target sea area to ensure the national maritime benefits, then monitor the quantity of the ship in the sea area and plan a reasonable channel. Because the ship distribution in the sea area is different, the investigation and research of the ship data in each sea area are required to be acquired, the ship density distribution state is required to be detected and analyzed in advance based on the GIS in the traditional method, AIS data information sent by the target ship in the sea area is directly collected, a model is built by using a statistical theory method, the ship density distribution state in the sea area can be rapidly analyzed, a large amount of resource consumption is saved, on one hand, the analysis result is used as the basis of the ship channel planning in the sea area, and on the other hand, the number of various ships in the sea area is rapidly acquired to complete the preliminary real-time monitoring task.
Disclosure of Invention
The ship density is a statistical characteristic reflecting the number of ships in a specific sea area, and can explore the actual distribution situation of the ships in the sea area, including the busy degree of the ships in the sea area and the strait, and the dangerous degree of sailing in the area. The invention aims to provide a sea area ship density nonparametric estimation method, which utilizes AIS real-time data to obtain the quantity of each sea area and strait ships at fixed intervals, adopts a kernel function nonparametric estimation method and empirical distribution approximate overall distribution to estimate the quantity of the ships, obtains a specific sea area ship distribution rule, and provides theoretical support for space-based remote sensing sea area monitoring and channel planning tasks.
The technical scheme of the invention is as follows: the sea area ship density non-parameter estimation method comprises the following specific steps:
step 1, acquiring corresponding ship numbers of different sizes and different types of ships in a specific sea area as samples through AIS information;
step 2, constructing a non-parameter estimation model taking a Gaussian function as a kernel function:
inputting samples of each category into a model to obtain the value of the optimal bandwidth h, thereby determining a non-parameter estimation model of the ship quantity of each category at the current moment, wherein the model is expressed in a probability density function and is different between the bandwidth and the samples;
step 3, constructing a sample distribution function approximate cumulative probability distribution function:
(1) Sequencing the number of ships from small to large to obtain a discrete group of data X= [ X ] 1 ,x 2 …x n ];
(2) Calculating the number value x of each ship k The corresponding probability density value:
(3) Accumulating the probability density values of all the previous samples to obtain a current sample distribution function value, and finally connecting the values of all the points to obtain an approximate cumulative distribution function
Step 4, randomly generating [0,1]]Is input into a sample distribution function to obtain the number p ofAn integer x k As an estimated value of the number of vessels:
and 5, updating the model by the supplementary data set, and obtaining the number of the ships in each category at the current moment according to the model to realize the real-time monitoring task of the ships in each sea area.
Further, in step 1, the corresponding number of vessels of different sizes and different types of vessels in a specific sea area is obtained as samples through AIS information, which is specifically as follows:
and respectively acquiring ship quantity information of each sea area: the ship has the length of 0-50 m, 50-100 m, 100-200 m, 200-300 m and more than 300 m, the ship is of the seven types of engineering ship, cargo ship, oil ship, tug, fishing ship, passenger ship and official law enforcement ship, the samples are collected once every 15 minutes and respectively classified into the sample sets, and the sample number can be updated in real time.
Further, the sea area comprises east sea, south sea, philippine sea, bus strait, sulu sea, suraweixi sea, and tin strait.
The beneficial effects of the invention are as follows: the method has the advantages that corresponding ship numbers of different sizes and different types of ships in a specific sea area are obtained through AIS information to serve as samples, a kernel function non-parameter estimation model is constructed to reflect the ship number distribution rule of the sea area, the information processing speed is faster than that of the ship distribution rule obtained after GIS ship detection and analysis, the model can be updated in real time according to the added samples, the model can reflect more accurate ship number distribution characteristics, the ship numbers of different sizes and different types can be estimated in real time according to requirements, and then support can be provided for parallel processing of a plurality of remote sensing monitoring tasks.
Drawings
FIG. 1 is a diagram of regional ship number information for each sea area;
FIG. 2 is a non-parametric estimation model of the number of vessels in each category at the current moment;
FIG. 3 is a schematic diagram of estimating in real time the number of vessels of different sizes and different types under a functional model;
FIG. 4 is a flow chart of a non-parametric estimation method using a Gaussian function as a kernel function.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
As shown in fig. 4, by using a non-parameter estimation method using a gaussian function as a kernel function, eight sea areas including east sea, south sea, philippine sea, bus strait, sunlu sea, sunwei sienna and wandering strait are used as monitored object areas, and simultaneously, the number of ships of different sizes and types in each area in the same period is obtained, a statistical model is built, and kernel function optimal bandwidth parameters of different areas are obtained, so that a non-parameter estimation model is obtained. The method comprises the following specific steps:
step 1, respectively acquiring ship quantity information of each sea area: the ship has the length of 0-50 m, 50-100 m, 100-200 m, 200-300 m and more than 300 m, and the ship is of seven types of engineering ship, cargo ship, oil ship, tug, fishing ship, passenger ship and official law enforcement ship, as shown in figure 1, the samples are collected once every 15 minutes and respectively classified into the sample sets of the category, and the sample number can be updated in real time.
Step 2, constructing a non-parameter estimation model taking a Gaussian function as a kernel function:
and inputting the samples of each category into a model to obtain the value of the optimal bandwidth h, so as to determine a non-parameter estimation model of the ship quantity of each category at the current moment, wherein the model is expressed in the form of a probability density function and is different between the bandwidth and the samples as shown in fig. 2.
Step 3, constructing a sample distribution function approximate cumulative probability distribution function:
(1) Sequencing the number of ships from small to large to obtain a discrete group of data X= [ X ] 1 ,x 2 …x n ]。
(2) Calculating the number value x of each ship k The corresponding probability density value:
(3) Accumulating the probability density values of all the previous samples to obtain a current sample distribution function value, and finally connecting the values of all the points to obtain an approximate cumulative distribution function
Step 4, randomly generating [0,1]]Is input into a sample distribution function to obtain an integer x k As an estimated value of the number of vessels:
and 5, updating the model by the supplementary data set, and obtaining the number of the ships in each category at the current moment according to the model to realize the real-time monitoring task of the ships in each sea area.
The kernel function non-parameter estimation method comprises the following steps: after the frequency histogram of the number of vessels is obtained, the abscissa represents the number of vessels, the ordinate represents the density of vessels, and the actual density of vessels needs to be estimated for each value of the number of vessels so as to be more fit with the distribution rule of vessels. And calculating a density value by using a kernel function according to the distance between the point to be estimated and each sample point (broadband h represents a distance attenuation threshold), wherein the closer the distance is, the larger the obtained density value is, the smaller the distance is, and finally, the probability density value of the estimated point in sample distribution is obtained by weighted average of all the density values, so that the probability estimation obtained by each ship quantity value is a weighted value obtained by jointly participating in calculation of all the sample points. The factors that determine the effect of the non-parametric estimation are the kernel function K (x) and the choice of bandwidth h and the sufficient number of samples n.
Let the ship probability density be:
the kernel function K is a primary factor for determining whether a fitting result is accurate, a Gaussian kernel function is selected based on the principles of nonnegativity, symmetry and normalization, standard normal distribution is obeyed, and a smooth kernel function is adopted for fitting observed data points, so that a real probability distribution curve is simulated. Another important factor bandwidth selection principle: as shown in fig. 2, the smaller the selected bandwidth value, the less smooth the probability density function is, and the more noise is contained. If the selected bandwidth value is too large, the generated probability density function is too smooth, and the less detail is contained. And determining a proper bandwidth value h by judging that the curve is smooth and accords with the curve detail and simultaneously adopting the maximum value of the ship quantity calculated by the bandwidth to accord with the maximum value after curve fitting.
And determining a non-parameter estimation model of each category, wherein the model is expressed in a probability density function shown in a formula (1), and a cumulative distribution function is obtained according to the probability density function through a formula (2), so that the number of ships in a certain category in the sea area at the current moment can be estimated according to an inverse function of the cumulative distribution function of the formula (7). The condition that the cumulative distribution function has an inverse function is a continuous strictly increasing function, and when the number of samples is sufficiently large, the sample distribution function well approximates the true cumulative distribution function.
Let F (X) =y, then y e (0, 1) represents the probability that the number of vessels is not greater than X.
x=F -1 (x)=L(y) (7)
The ship quantity probability density function of each sea area is obtained by using a kernel function estimation method, the function reflects the ship quantity distribution rule in the long term of the corresponding sea area, a more data real-time update function form (mainly updating bandwidth value) can be added, the ship quantity of the sea area can be estimated in real time by using the function, the corresponding ship quantity of different sizes and different types of ships in a certain sea area can be obtained through AIS information, the quantity of the corresponding types can be calculated under the function model, and the ship quantity of different sizes and different types can be estimated in real time, as shown in figure 3. And providing statistical data support for remote sensing monitoring tasks by using a sea area ship quantity non-parameter estimation model.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention.
Claims (3)
1. The sea area ship density non-parameter estimation method is characterized by comprising the following specific steps:
step 1, acquiring corresponding ship numbers of different sizes and different types of ships in a specific sea area as samples through AIS information;
step 2, constructing a non-parameter estimation model taking a Gaussian function as a kernel function:
inputting samples of each category into a model to obtain the value of the optimal bandwidth h, thereby determining a non-parameter estimation model of the ship quantity of each category at the current moment, wherein the model is expressed in a probability density function and is different between the bandwidth and the samples;
step 3, constructing a sample distribution function approximate cumulative probability distribution function:
(1) Sequencing the number of ships from small to large to obtain a discrete group of data X= [ X ] 1 ,x 2 …x n ];
(2) Calculating the number value x of each ship k The corresponding probability density value:
(3) Accumulating the probability density values of all the previous samples to obtain the current sample distribution function value, and finally connecting the values of all the pointsObtaining an approximate cumulative distribution function
Step 4, randomly generating [0,1]]Is input into a sample distribution function to obtain an integer x k As an estimated value of the number of vessels:
and 5, updating the model by the supplementary data set, and obtaining the number of the ships in each category at the current moment according to the model to realize the real-time monitoring task of the ships in each sea area.
2. The sea area ship density nonparametric estimation method according to claim 1, characterized in that: in the step 1, the corresponding number of ships with different sizes and different types in a specific sea area is obtained as samples through AIS information, and the specific steps are as follows:
and respectively acquiring ship quantity information of each sea area: the ship has the length of 0-50 m, 50-100 m, 100-200 m, 200-300 m and more than 300 m, the ship is of the seven types of engineering ship, cargo ship, oil ship, tug, fishing ship, passenger ship and official law enforcement ship, the samples are collected once every 15 minutes and respectively classified into the sample sets, and the sample number can be updated in real time.
3. The sea area ship density nonparametric estimation method according to claim 1, characterized in that: the sea area comprises east sea, south sea, philippine sea, bus strait, sulu sea, suravigneaux sea, and tin strait.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311282157.3A CN117688288A (en) | 2023-10-07 | 2023-10-07 | Sea area ship density non-parameter estimation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311282157.3A CN117688288A (en) | 2023-10-07 | 2023-10-07 | Sea area ship density non-parameter estimation method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117688288A true CN117688288A (en) | 2024-03-12 |
Family
ID=90127298
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311282157.3A Pending CN117688288A (en) | 2023-10-07 | 2023-10-07 | Sea area ship density non-parameter estimation method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117688288A (en) |
-
2023
- 2023-10-07 CN CN202311282157.3A patent/CN117688288A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110415516B (en) | Urban traffic flow prediction method and medium based on graph convolution neural network | |
CN110210508B (en) | Model generation method, abnormal flow detection device, electronic device and computer-readable storage medium | |
CN111222526B (en) | Method, device, equipment and storage medium for identifying real-time fishing behavior of fishing vessel | |
CN112883648B (en) | Training method and device for automobile fuel consumption prediction model and computer equipment | |
CN109299816A (en) | Ship track real-time predicting method based on Rolling Planning strategy | |
CN111814956A (en) | Multi-task learning air quality prediction method based on multi-dimensional secondary feature extraction | |
CN110473085B (en) | Vehicle risk discrimination method and device | |
CN114493680B (en) | Fishery resource statistical method and system based on stream stab net investigation | |
CN112949932B (en) | Ship traffic flow prediction method, device, computer equipment and storage medium | |
CN113269805B (en) | Rainfall event guided remote sensing rainfall inversion training sample self-adaptive selection method | |
CN111931991A (en) | Weather nowcast method, weather nowcast device, computer equipment and storage medium | |
CN115221973A (en) | Aviation bearing fault diagnosis method based on enhanced weighted heterogeneous ensemble learning | |
CN113032239A (en) | Risk prompting method and device, electronic equipment and storage medium | |
CN117688288A (en) | Sea area ship density non-parameter estimation method | |
CN117421994A (en) | Edge application health monitoring method and system | |
CN112733935A (en) | Fishing boat fishing mode prediction method based on Stacking algorithm | |
CN115907079B (en) | Airspace traffic flow prediction method based on attention space-time diagram convolutional network | |
CN108304610B (en) | Dynamic tracking method for atmosphere high-pollution process | |
CN116721113A (en) | Three-dimensional point cloud plane segmentation method and system | |
CN112085926B (en) | River water pollution early warning method and system | |
CN114492540A (en) | Training method and device of target detection model, computer equipment and storage medium | |
CN114565004A (en) | Method and device for eliminating abnormal scattered points of power curve of wind turbine generator | |
CN111583655A (en) | Traffic flow detection method, device, equipment and medium | |
CN112565346A (en) | Internet of vehicles data acquisition method and device, electronic equipment and readable storage medium | |
CN110991685A (en) | Meteorological temperature prediction method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |