CN113112059A - Ship berthing time prediction method and system - Google Patents
Ship berthing time prediction method and system Download PDFInfo
- Publication number
- CN113112059A CN113112059A CN202110349387.1A CN202110349387A CN113112059A CN 113112059 A CN113112059 A CN 113112059A CN 202110349387 A CN202110349387 A CN 202110349387A CN 113112059 A CN113112059 A CN 113112059A
- Authority
- CN
- China
- Prior art keywords
- berthing time
- target
- ship information
- ship
- target ship
- 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 51
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims description 15
- 230000006870 function Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 9
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 claims description 4
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 101150080778 INPP5D gene Proteins 0.000 description 137
- 238000004891 communication Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Marketing (AREA)
- Biophysics (AREA)
- General Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method and a system for predicting ship berthing time, wherein the method for predicting the ship berthing time comprises the following steps: acquiring target ship information; inputting the target ship information into a berthing time prediction model, and outputting target berthing time; the berthing time prediction model is obtained by training by taking sample ship information as a sample and taking sample berthing time corresponding to the sample ship information as a sample label. According to the ship berthing time prediction method and system, the trained berthing time prediction model is used for predicting the target berthing time according to the target ship information, so that the actual value can be more approximate, and the prediction accuracy of the ship berthing time is improved.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a method and a system for predicting berthing time of a ship.
Background
For a marine vessel taking a container liner as an example, a schedule is an important basis for a cargo owner to order a tank and make a logistics transportation plan, and a schedule is an important basis for the vessel to make a berth.
At present, in order to predict the berthing time of a ship, operation and maintenance companies corresponding to the ship are often inquired through agents, but the mode is low in efficiency and only approximate time can be obtained, and methods for tracking the sailing track of the ship and real-time dynamic states through a satellite AIS technology are also available.
Disclosure of Invention
The invention provides a method for predicting ship berthing time, which is used for overcoming the defect of low accuracy of the predicted ship berthing time in the prior art and realizing the improvement of the prediction accuracy of the ship berthing time.
The invention provides a method for predicting the berthing time of a ship, which comprises the following steps: acquiring target ship information; inputting the target ship information into a berthing time prediction model, and outputting target berthing time; the berthing time prediction model is obtained by training by taking sample ship information as a sample and taking sample berthing time corresponding to the sample ship information as a sample label.
According to the ship berthing time prediction method provided by the invention, the berthing time prediction model comprises the following steps: a Wide submodel, a Deep submodel, an LSTM submodel and a regression prediction layer; inputting the target ship information into a berthing time prediction model and outputting a target berthing time, wherein the berthing time prediction model comprises the following steps: inputting the target ship information into the Wide sub-model and outputting first reference data; inputting the target ship information into the Deep submodel, and outputting second reference data; inputting the target ship information into the LSTM submodel, and outputting third reference data; inputting the first reference data, the second reference data and the third reference data into the regression prediction layer, and outputting the target berthing time.
According to the method for predicting the berthing time of the ship, provided by the invention, the Wide sub-model comprises the following steps: the system comprises a first embedding layer used for operating class characteristics in the target ship information, a first containing layer used for connecting the class characteristics and numerical characteristics in the target ship information, a cross product layer used for performing characteristic cross conversion, and a second containing layer used for connecting the converted characteristics and original characteristics in the target ship information; the Deep submodel comprises: the first embedding layer shared with the Wide submodel, the first concatenate layer shared with the Wide submodel and a 4-layer full-connection layer network with the activation function ReLu and the parameters of dropout of 0.5, 0.5 and 0.3 respectively; the LSTM submodel comprises: the system comprises a second embedding layer for operating the class characteristics in the target ship information, a third containing layer for connecting the class characteristics and the numerical characteristics in the target ship information, two LSTM layers and a full-connection layer network with an activation function ReLu and a parameter of dropout being 0.1.
According to the ship berthing time prediction method provided by the invention, the target ship information is input into a berthing time prediction model, and the target berthing time is output, and the method comprises the following steps: preprocessing the target ship information to obtain target ship tensor data; and inputting the tensor data of the target ship into the berthing time prediction model, and outputting the berthing time of the target ship.
According to the method for predicting the berthing time of the ship, provided by the invention, the target ship information is preprocessed to obtain tensor data of the target ship, and the method comprises the following steps: and at least one of encoding processing, normalization processing and sequencing processing is carried out on the target ship information to obtain tensor data of the target ship.
According to the method for predicting the berthing time of the ship provided by the invention, the target ship information comprises: at least one of a current voyage characteristic, a historical statistical characteristic, a vessel base characteristic, a destination port characteristic, and a queued vessel aggregate characteristic.
According to the ship berthing time prediction method provided by the invention, the berthing time prediction model is obtained by training based on a back propagation algorithm and a gradient descent algorithm by taking MAPE as a loss function.
The invention also provides a system for predicting the berthing time of a ship, which comprises: the acquisition module is used for acquiring target ship information; the output module is used for inputting the target ship information into a berthing time prediction model and outputting target berthing time; the berthing time prediction model is obtained by training by taking sample ship information as a sample and taking sample berthing time corresponding to the sample ship information as a sample label.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program to realize the steps of any one of the ship berthing time prediction methods.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, carries out the steps of the method of predicting a berthing time of a vessel as claimed in any one of the above.
According to the ship berthing time prediction method and system, the trained berthing time prediction model is used for predicting the target berthing time according to the target ship information, so that the actual value can be more approximate, and the prediction accuracy of the ship berthing time is improved.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a method for predicting a berthing time of a ship according to the present invention;
FIG. 2 is a schematic structural diagram of a berthing time prediction model of the ship berthing time prediction method according to the present invention;
FIG. 3 is a second schematic structural diagram of a berthing time prediction model of the method for predicting berthing time of ships according to the present invention;
FIG. 4 is a schematic structural diagram of a system for predicting berthing time of a ship according to the present invention;
fig. 5 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The method and system for predicting the berthing time of a ship according to the present invention will be described with reference to fig. 1 to 5.
As shown in fig. 1, the present invention provides a method for predicting a berthing time of a ship, which includes the following steps 110 to 120.
And step 110, acquiring target ship information.
It can be understood that the target ship is an object for which the berthing time of the ship needs to be predicted, and the target ship information is a characteristic related to the ship navigation, such as a basic characteristic which may include the name of the ship, the tonnage of the ship, the heading of the ship, and the speed of the ship, and can reflect the physical parameters of the ship navigation.
In some embodiments, the target vessel information includes: at least one of a current voyage characteristic, a historical statistical characteristic, a vessel base characteristic, a destination port characteristic, and a queued vessel aggregate characteristic.
It will be appreciated that the set of queued vessels is the set of all other underway vessels that are coincident with and closer to the destination dock of the target vessel.
The current voyage features include: the previous port ID, the next port ID, the destination port ID, the current latitude and longitude, the area where the ship is located, the weather characteristics of the area where the ship is located, the current speed, and the next port ETA (Estimated Time of Arrival) on the ship stage, etc., are used to characterize the current state of the target ship or the queued ship.
The historical statistical characteristics include: the method comprises the steps that all historical affiliated operation time, waiting time and ship-stage error statistical values of the ship are obtained, wherein the statistical indexes are selected from a mean value, a standard deviation, a 20% quantile and an 80% quantile, and the statistical values are updated once every week and are used for representing the operation efficiency and the shift-ready rate of a target ship or a queued ship.
The basic features of the ship include: and the ship name, the ship length, the ship company, the load capacity ton, the maximum draft and other basic information are used for representing the basic properties of the target ship or the queued ship.
The purpose wharf characteristic includes: the basic characteristics of the wharf, the current time characteristics of the wharf and historical statistical characteristics of the wharf. The basic characteristics of the wharf comprise basic information such as wharf ID and wharf length; the current time characteristics of the wharf comprise the number of ships operating on the wharf, the total length of the ships, the description of the remaining berths of the wharf, the number of the ships with the current wharf serving as a target wharf on an anchor ground, the total length of the ships and the like, and are used for representing the current busy degree of the wharf; the historical statistical characteristics of the wharf comprise statistical values of all hanging operation time, waiting time and ship-term errors of the wharf in history, wherein the statistical indexes are selected from a mean value, a standard deviation, a 20% quantile and an 80% quantile, and the statistical values are updated once per week and are used for representing the operation efficiency and the shift-ready rate of the wharf.
The destination port characteristics include: the current time characteristic of the port and the historical statistical characteristic of the port. The current time characteristics of the port comprise the number of ships in the port area, the total length of the ships, meteorological characteristics of the port area and the like, and are used for representing the current busy degree and meteorological state of the port; the historical statistical characteristics of the port comprise statistical values of all hanging operation time, waiting time and ship-term errors in the port history, wherein the statistical indexes are selected from a mean value, a standard deviation, a 20% quantile and an 80% quantile, and the statistical values are updated once per week and are used for representing the operation efficiency and the shift-ready rate of the target port.
The queued vessel set characteristics are a current voyage characteristic of each queued vessel within the queued vessel set, historical statistical characteristics of each queued vessel, and vessel base characteristics of each queued vessel.
And 120, inputting the target ship information into the berthing time prediction model, and outputting the target berthing time.
It can be understood that the berthing time prediction model can be a neural network model, a large number of ship samples can be adopted to train the neural network model, and the trained berthing time prediction model can be used for efficiently and accurately predicting the target berthing time based on target ship information.
The berthing time prediction model is obtained by training by taking the sample ship information as a sample and taking the sample berthing time corresponding to the sample ship information as a sample label.
It will be appreciated that the sample vessel may be all container vessels at various times and for different purposes docks that have generated actual berthing times. The sample vessel information may also include: at least one of a current voyage characteristic, a historical statistical characteristic, a vessel base characteristic, a destination port characteristic, and a queued vessel aggregate characteristic.
The sample set used for training the berthing time prediction model is provided with a large amount of sample ship information and sample berthing time corresponding to the sample ship information, and the corresponding relation between the sample berthing time and the sample ship information can reflect that the berthing time is influenced by various natural factors or human factors in ship navigation under the real condition.
The trained berthing time prediction model can be applied to the berthing time prediction of the target ship, and after target ship information is input into the berthing time prediction model, the berthing time of the target ship can be predicted efficiently and conveniently.
According to the ship berthing time prediction method provided by the invention, the trained berthing time prediction model is used for predicting the target berthing time according to the target ship information, so that the actual value can be more approximate, and the prediction accuracy of the ship berthing time is improved.
In some embodiments, inputting the target vessel information into a berthing time prediction model, outputting a target berthing time, comprises: inputting target ship information into the Wide sub-model, and outputting first reference data; inputting the target ship information into a Deep submodel, and outputting second reference data; inputting the target ship information into an LSTM submodel, and outputting third reference data; and inputting the first reference data, the second reference data and the third reference data into a regression prediction layer, and outputting the target berthing time.
It can be understood that the structure of the berthing time prediction model provided by this embodiment is constructed based on the ideas of Wide & Deep network and LSTM neurons, and thus may also be referred to as a WDL model, and the berthing time prediction model has three sub-network blocks and a regression prediction layer, and the three sub-network blocks are respectively a Wide sub-model, a Deep sub-model and an LSTM sub-model.
As shown in FIG. 2, the Wide & Deep network is a class of models which are issued by Google and applied to classification and regression, and the core idea is that the parameters of 2 models are optimized simultaneously in the training process by combining the memory capacity of a linear model and the generalization capacity of a DNN model, so that the optimal prediction capacity of the whole model is achieved.
The Wide sub-model is of the form y ═ wTThe generalized linear model of x + b, structure is shown in the left half of FIG. 2.
Wherein y is a predicted value, and x ═ x1，x2，...，xd]Is a vector containing d features, w ═ w1，w2，...，wd]Is a modelThe parameter, T is the matrix transpose operator, and b is the offset. The feature set comprises original input features and converted features, and the conversion is cross product conversion defined as:
wherein k denotes the subscript of the kth cross feature, ckiThe power exponent of the ith original feature at the time of the kth cross feature generation.
The Deep sub-model is a feedforward neural network model, and the structure is shown in the right half of fig. 2. The Deep model firstly maps the class characteristics into numerical characteristics through a characteristic embedding layer, and then is connected with the numerical characteristics of the original input into a characteristic vector to be used as the input of a neural network hiding layer. The hidden layer performs the following operation a(l+1)＝f(W(l)a(l)+b(l)) Where l is the layer number and f is the activation function, usually a corrective linear unit function (ReLUs) is used. a is(l)，b(l)And w(l)The activation value, bias and weight of the l-th layer of the model, respectively.
The LSTM submodel belongs to a recurrent neural network and can capture local features of each time step and long-term dependence features of a sequence at the same time. The LSTM submodel uses several gates to control the information flow. In each inference step of the LSTM submodel, the input gate, the forgetting gate, the output gate and the modulation input are updated in the following way, respectively:
it＝σ(Wi[xt；ht-1]+bi)，
ft＝σ(Wf[xt；ht-1]+bf)，
ot＝σ(Wo[xt；ht-1]+bo)，
gt＝σ(Wg[xt；ht-1]+bg)，
wherein, Wi，Wf，Wo，WgUpdate weights for input gate, forgetting gate, output gate and modulation input, respectively, bi，bf，bo，bgRespectively, update the offset, xtFor input of the current time step, ht-1Is the hidden layer state of the last time step. σ () is sigmoid function σ (u) 1/(1+ e)-u). Each equation contains an affine transformation and a non-linear activation. The state of the memory neuron and the hidden layer is updated as follows:
ct＝ft⊙ct-1+it⊙gt，
ht＝ot⊙tanh(ct)，
wherein, l is an element multiplication, tanh is a hyperbolic tangent function, ft，it，gtAnd otA forgetting gate state, an input gate state, an output gate state and a modulation input state of the current time step, respectively, ct-1The state of the memory neuron at the last time step.
The method comprises the steps that first reference data are obtained by inputting target ship information into a Wide submodel, second reference data are obtained by inputting the target ship information into a Deep submodel, third reference data are obtained by inputting the target ship information into an LSTM submodel, the first reference data, the second reference data and the third reference data are data in an intermediate form, finally the first reference data, the second reference data and the third reference data are connected in a regression prediction layer, and the regression prediction layer outputs target berthing time.
In some embodiments, the berthing time prediction model is obtained by training based on a back propagation algorithm and a gradient descent algorithm by taking MAPE as a loss function.
In other words, when the berthing time prediction model is trained, the sample ship information and the corresponding berthing time of the sample can be used as a training data set, and under the condition that the MAPE is used as a loss function, all parameters in the WDL model are jointly trained through a back propagation algorithm and a gradient descent algorithm.
Since the combination of 3 sub-network blocks in the berthing time prediction model makes the global learning rate difficult to obtain, a random gradient descent method (Adam) with adaptive step size and momentum can be selected to optimize the berthing time prediction model.
As shown in fig. 3, in some embodiments, the Wide sub-model includes: the system comprises a first embedding layer for operating the class characteristics in the target ship information, a first containing layer for connecting the class characteristics and the numerical characteristics in the target ship information, a cross product layer for performing characteristic cross conversion, and a second containing layer for connecting the converted characteristics and the original characteristics in the target ship information.
The Wide submodel may also have a layer that performs feature cross-conversion.
The Deep submodel comprises: the first embedding layer shared with the Wide submodel, the first concatenate layer shared with the Wide submodel and the 4-layer full-link layer network with the activation function ReLu and the dropout with the parameters of 0.5, 0.5 and 0.3 respectively.
The LSTM submodel includes: the system comprises a second embedding layer for operating the class characteristics in the target ship information, a third containing layer for connecting the class characteristics and the numerical characteristics in the target ship information, two LSTM layers and a full-connection layer network with the activation function ReLu and the parameter of dropout being 0.1.
In some embodiments, step 120, inputting the target ship information into a berthing time prediction model, and outputting the target berthing time includes: preprocessing the target ship information to obtain target ship tensor data; and inputting the tensor data of the target ship into a berthing time prediction model, and outputting the berthing time of the target ship.
It can be understood that before the target ship information is input into the berthing time prediction model, the target ship information can be preprocessed to obtain target ship tensor data, and the target ship tensor data is data in a tensor form suitable for being input by a neural network and can be convenient for the berthing time prediction model to process.
In some embodiments, preprocessing the target vessel information to obtain target vessel tensor data includes: and at least one of encoding processing, normalization processing and sequencing processing is carried out on the target ship information to obtain target ship tensor data.
It is to be understood that the preprocessing may be at least one of encoding processing, normalization processing, and sorting processing, the encoding processing is for data of a category type, the normalization processing is for data of a numerical type, and the sorting processing is for generating the queued ship sequence data.
Notably, the preprocessing of the model application process at the berthing time prediction model is logically consistent with the preprocessing of the model training process.
The following describes the ship berthing time prediction system provided by the present invention, and the ship berthing time prediction system described below and the ship berthing time prediction method described above may be referred to correspondingly.
As shown in fig. 4, the present invention also provides a ship berthing time prediction system, including: an acquisition module 410 and an output module 420.
An obtaining module 410, configured to obtain target ship information.
The output module 420 is used for inputting the target ship information into the berthing time prediction model and outputting the target berthing time; the berthing time prediction model is obtained by training by taking the sample ship information as a sample and taking the sample berthing time corresponding to the sample ship information as a sample label.
Fig. 5 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 5: a processor (processor)510, a communication Interface (Communications Interface)520, a memory (memory)530 and a communication bus 540, wherein the processor 510, the communication Interface 520 and the memory 530 communicate with each other via the communication bus 540. Processor 510 may invoke logic instructions in memory 530 to perform a method of vessel berthing time prediction, the method comprising: acquiring target ship information; inputting target ship information into a berthing time prediction model, and outputting target berthing time; the berthing time prediction model is obtained by training by taking the sample ship information as a sample and taking the sample berthing time corresponding to the sample ship information as a sample label.
Furthermore, the logic instructions in the memory 530 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform a method for predicting a berthing time of a ship provided by the above methods, the method comprising: acquiring target ship information; inputting target ship information into a berthing time prediction model, and outputting target berthing time; the berthing time prediction model is obtained by training by taking the sample ship information as a sample and taking the sample berthing time corresponding to the sample ship information as a sample label.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the ship berthing time prediction methods provided above, the method comprising: acquiring target ship information; inputting target ship information into a berthing time prediction model, and outputting target berthing time; the berthing time prediction model is obtained by training by taking the sample ship information as a sample and taking the sample berthing time corresponding to the sample ship information as a sample label.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for predicting a berthing time of a ship, comprising:
acquiring target ship information;
inputting the target ship information into a berthing time prediction model, and outputting target berthing time;
the berthing time prediction model is obtained by training by taking sample ship information as a sample and taking sample berthing time corresponding to the sample ship information as a sample label.
2. The ship berthing time prediction method of claim 1, wherein the berthing time prediction model comprises: a Wide submodel, a Deep submodel, an LSTM submodel and a regression prediction layer;
inputting the target ship information into a berthing time prediction model and outputting a target berthing time, wherein the berthing time prediction model comprises the following steps:
inputting the target ship information into the Wide sub-model and outputting first reference data;
inputting the target ship information into the Deep submodel, and outputting second reference data;
inputting the target ship information into the LSTM submodel, and outputting third reference data;
inputting the first reference data, the second reference data and the third reference data into the regression prediction layer, and outputting the target berthing time.
3. The method of predicting berthing time of a vessel according to claim 2,
the Wide sub-model comprises the following steps: the system comprises a first embedding layer used for operating class characteristics in the target ship information, a first containing layer used for connecting the class characteristics and numerical characteristics in the target ship information, a cross product layer used for performing characteristic cross conversion, and a second containing layer used for connecting the converted characteristics and original characteristics in the target ship information;
the Deep submodel comprises: the first embedding layer shared with the Wide submodel, the first concatenate layer shared with the Wide submodel and a 4-layer full-connection layer network with the activation function ReLu and the parameters of dropout of 0.5, 0.5 and 0.3 respectively;
the LSTM submodel comprises: the system comprises a second embedding layer for operating the class characteristics in the target ship information, a third containing layer for connecting the class characteristics and the numerical characteristics in the target ship information, two LSTM layers and a full-connection layer network with an activation function ReLu and a parameter of dropout being 0.1.
4. The method of claim 1, wherein the inputting the target ship information into a berthing time prediction model and outputting a target berthing time comprises:
preprocessing the target ship information to obtain target ship tensor data;
and inputting the tensor data of the target ship into the berthing time prediction model, and outputting the berthing time of the target ship.
5. The method for predicting the berthing time of a ship according to claim 4, wherein the preprocessing the target ship information to obtain target ship tensor data comprises:
and at least one of encoding processing, normalization processing and sequencing processing is carried out on the target ship information to obtain tensor data of the target ship.
6. The method according to any one of claims 1 to 5, wherein the target vessel information includes: at least one of a current voyage characteristic, a historical statistical characteristic, a vessel base characteristic, a destination port characteristic, and a queued vessel aggregate characteristic.
7. The method according to any one of claims 1 to 5, wherein the berthing time prediction model is obtained by training based on a back propagation algorithm and a gradient descent algorithm with MAPE as a loss function.
8. A system for predicting a berthing time of a ship, comprising:
the acquisition module is used for acquiring target ship information;
the output module is used for inputting the target ship information into a berthing time prediction model and outputting target berthing time;
the berthing time prediction model is obtained by training by taking sample ship information as a sample and taking sample berthing time corresponding to the sample ship information as a sample label.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the method of predicting a berthing time of a vessel according to any of claims 1 to 7.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the steps of the method for predicting a berthing time of a vessel according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110349387.1A CN113112059A (en) | 2021-03-31 | 2021-03-31 | Ship berthing time prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110349387.1A CN113112059A (en) | 2021-03-31 | 2021-03-31 | Ship berthing time prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113112059A true CN113112059A (en) | 2021-07-13 |
Family
ID=76713252
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110349387.1A Pending CN113112059A (en) | 2021-03-31 | 2021-03-31 | Ship berthing time prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113112059A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115410417A (en) * | 2022-07-20 | 2022-11-29 | 国能浙江北仑第一发电有限公司 | Ship dispatching berthing prediction system based on environmental factors |
| CN115563889A (en) * | 2022-12-06 | 2023-01-03 | 三亚海兰寰宇海洋信息科技有限公司 | Target object anchoring prediction method, device and equipment |
| CN116911468A (en) * | 2023-09-12 | 2023-10-20 | 中国标准化研究院 | Global shipping information monitoring system based on big data |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109002905A (en) * | 2018-06-25 | 2018-12-14 | 北京嘀嘀无限科技发展有限公司 | Estimate the method and system of arrival time |
| CN109196527A (en) * | 2016-04-13 | 2019-01-11 | 谷歌有限责任公司 | Breadth and depth machine learning model |
| CN109416878A (en) * | 2017-06-13 | 2019-03-01 | 北京嘀嘀无限科技发展有限公司 | System and method for recommending E.T.A |
| CN110073426A (en) * | 2017-11-23 | 2019-07-30 | 北京嘀嘀无限科技发展有限公司 | The system and method for Estimated Time of Arrival |
| CN111461426A (en) * | 2020-03-30 | 2020-07-28 | 南京邮电大学 | High-precision travel time length prediction method based on deep learning |
| CN111563072A (en) * | 2020-04-15 | 2020-08-21 | 交通运输部水运科学研究所 | AIS information-based ship real-time accurate position acquisition method |
| CN111832881A (en) * | 2020-04-30 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Method, medium and electronic device for predicting electric vehicle energy consumption based on road condition information |
-
2021
- 2021-03-31 CN CN202110349387.1A patent/CN113112059A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109196527A (en) * | 2016-04-13 | 2019-01-11 | 谷歌有限责任公司 | Breadth and depth machine learning model |
| CN109416878A (en) * | 2017-06-13 | 2019-03-01 | 北京嘀嘀无限科技发展有限公司 | System and method for recommending E.T.A |
| US20200011692A1 (en) * | 2017-06-13 | 2020-01-09 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for recommending an estimated time of arrival |
| CN110073426A (en) * | 2017-11-23 | 2019-07-30 | 北京嘀嘀无限科技发展有限公司 | The system and method for Estimated Time of Arrival |
| CN109002905A (en) * | 2018-06-25 | 2018-12-14 | 北京嘀嘀无限科技发展有限公司 | Estimate the method and system of arrival time |
| WO2020001261A1 (en) * | 2018-06-25 | 2020-01-02 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for estimating an arriel time of a vehicle |
| CN111461426A (en) * | 2020-03-30 | 2020-07-28 | 南京邮电大学 | High-precision travel time length prediction method based on deep learning |
| CN111563072A (en) * | 2020-04-15 | 2020-08-21 | 交通运输部水运科学研究所 | AIS information-based ship real-time accurate position acquisition method |
| CN111832881A (en) * | 2020-04-30 | 2020-10-27 | 北京嘀嘀无限科技发展有限公司 | Method, medium and electronic device for predicting electric vehicle energy consumption based on road condition information |
Non-Patent Citations (1)
| Title |
|---|
| 吕海灿;王伟峰;赵兵;张毅;郭秋婷;胡伟;: "基于Wide&Deep-LSTM模型的短期台区负荷预测", 电网技术, no. 02 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115410417A (en) * | 2022-07-20 | 2022-11-29 | 国能浙江北仑第一发电有限公司 | Ship dispatching berthing prediction system based on environmental factors |
| CN115410417B (en) * | 2022-07-20 | 2023-11-14 | 国能浙江北仑第一发电有限公司 | Ship scheduling berthing prediction system based on environmental factors |
| CN115563889A (en) * | 2022-12-06 | 2023-01-03 | 三亚海兰寰宇海洋信息科技有限公司 | Target object anchoring prediction method, device and equipment |
| CN116911468A (en) * | 2023-09-12 | 2023-10-20 | 中国标准化研究院 | Global shipping information monitoring system based on big data |
| CN116911468B (en) * | 2023-09-12 | 2023-12-08 | 中国标准化研究院 | Global shipping information monitoring system based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113112059A (en) | Ship berthing time prediction method and system | |
| Barua et al. | Machine learning for international freight transportation management: A comprehensive review | |
| Zhong et al. | Inland ship trajectory restoration by recurrent neural network | |
| Kawai et al. | Sea state estimation using monitoring data by convolutional neural network (CNN) | |
| Huang et al. | Machine learning in sustainable ship design and operation: A review | |
| Le Carrer et al. | Optimising cargo loading and ship scheduling in tidal areas | |
| Gao et al. | The daily container volumes prediction of storage yard in port with long short-term memory recurrent neural network | |
| Chaal | Ship operational performance modelling for voyage optimization through fuel consumption minimization | |
| Maiorov et al. | Improvement of the quality of the sea passenger terminal based on methods of forecasting | |
| Mittendorf et al. | The prediction of sea state parameters by deep learning techniques using ship motion data | |
| Ultsch et al. | Self-organizing feature maps predicting sea levels | |
| Rabar et al. | Dry-docking performance measurement model–multi criteria non parametric approach | |
| CN115410419B (en) | Ship mooring early warning method, system, electronic equipment and storage medium | |
| Gaidai et al. | Gaidai multivariate risk assessment method for cargo ship dynamics | |
| CN111797474A (en) | Method, device and medium for establishing and predicting ship dynamic stability performance prediction model | |
| Mei et al. | Ship maneuvering prediction using grey box framework via adaptive RM-SVM with minor rudder | |
| Wróbel et al. | With regard to the autonomy in maritime operations–hydrography and shipping, interlinked | |
| Zhou et al. | Estimation of shipment size in seaborne iron ore trade | |
| Petersen | Mining of ship operation data for energy conservation | |
| Uğurlu et al. | An Awesim simulation study: To determine the efficiency of future improvements on Tupras Izmit Oil Terminal | |
| Sun et al. | Deep learning-based trajectory tracking control forunmanned surface vehicle | |
| Bondarenko et al. | Effectiveness and optimization of harbour tug fleet | |
| Wang et al. | A system for container terminal operation prediction | |
| Eisen et al. | An Evaluation of First-Come, First-Served Scheduling in a Geometrically-Constrained Wet Bulk Terminal | |
| Krata et al. | The Development of a Combined Method to Quickly Assess Ship Speed and Fuel Consumption at Different Powertrain Load and Sea Conditions |
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 |