CN111523641A - ConvLSTM-SRU-based sector delay prediction method - Google Patents
ConvLSTM-SRU-based sector delay prediction method Download PDFInfo
- Publication number
- CN111523641A CN111523641A CN202010277615.4A CN202010277615A CN111523641A CN 111523641 A CN111523641 A CN 111523641A CN 202010277615 A CN202010277615 A CN 202010277615A CN 111523641 A CN111523641 A CN 111523641A
- Authority
- CN
- China
- Prior art keywords
- time
- sector
- flight
- network
- data
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000000605 extraction Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 4
- 238000010606 normalization Methods 0.000 claims description 24
- 238000012549 training Methods 0.000 claims description 23
- 238000010276 construction Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 13
- 230000004913 activation Effects 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 9
- 230000008859 change Effects 0.000 claims description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 4
- 210000002569 neuron Anatomy 0.000 claims description 4
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 claims description 3
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 208000027744 congestion Diseases 0.000 abstract 1
- 230000006870 function Effects 0.000 description 10
- 238000013135 deep learning Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
Images
Classifications
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield 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
- G06N3/04—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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G06Q50/40—
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention discloses a ConvLSTM-SRU-based sector delay prediction method, which comprises the following steps of: (1) reading historical flight ADS-B data of a sector; (2) processing data; (3) dividing index characteristics; (4) generating time-series data of each index; (5) aiming at the space-time characteristic index, establishing a ConvLSTM network space-time characteristic extraction model; (6) aiming at the time characteristic index, establishing an SRU network time characteristic extraction model; (7) establishing an index characteristic full-connection network model; (8) outputting a predicted sector flight time; (9) the expected delay time of the sector is calculated. The prediction method can improve the prediction precision of delay time of each route of the sector, provide forward-looking information for air traffic control personnel, help the control personnel to take control measures in advance, reduce the probability of increased congestion risk of the sector and further improve the operation efficiency of an airport.
Description
Technical Field
The invention belongs to the technical field of air traffic flow management, and particularly relates to a ConvLSTM-SRU-based sector delay prediction method.
Background
The sector is taken as a control airspace unit, aircrafts in the range are concentrated on each flight path, congestion is easy to occur under the condition of large traffic volume, and the congestion is spread to the whole sector and other sectors by the flight path. If the accuracy of predicting the delay time of each route of the sector in real time can be improved, important help is provided for relieving the congestion of the sector. The traditional methods for predicting the air traffic delay mostly judge according to the flow of the sector, and although the prediction methods can judge the route on which the delay exists, the prediction precision of the delay time is not enough, and the influence among the connected routes is not considered enough. In the process of sector operation, the aeronautical network structure of a sector is a key factor for determining the sector capacity, the flight time of an aircraft on a sector airline (from a sector entering point to a sector leaving point) is closely related to the operation state of the airline where the aircraft is located, and therefore two aspects of airline structure characteristics and flow quantity must be considered in the process of predicting delay time.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to provide a CNN-SRU-based sector delay time prediction method, which can effectively reduce the probability of sector congestion and improve the operation efficiency.
The technical scheme is as follows: the invention relates to a ConvLSTM-SRU-based sector delay prediction method, which comprises the following steps of:
(1) reading data: reading historical flight ADS-B data of the sector, counting to obtain flight flow data of each waypoint entering the sector and leaving the sector, flight flow data and flight time data of each airline (from the sector entering point to the sector leaving point) in the sector, and reading historical meteorological data of the sector in a corresponding time range;
(2) processing data: arranging the data sets read in the step (1) according to time tags, filling missing data on a time axis by using the mean values of the time periods before and after, and replacing abnormal data on the time axis by using an interpolation method;
(3) dividing index characteristics: the method comprises the following steps of performing type division on indexes according to index attributes, and dividing the two indexes into space-time characteristic indexes as the flight flow and the flight time of each route (from a sector entering point to a sector leaving point) in a sector change along with time and space (a sector route network); the flight flow and the meteorological conditions of each waypoint entering and leaving the sector only change along with time in the same sector and are irrelevant to the space, so the two indexes are divided into time characteristic indexes;
(4) generating time-series data of each index: on the basis of the step (3), selecting different time intervals (10min, 20min, 30min and 60min), and carrying out statistics on data of each index to generate sequence data of corresponding time intervals;
(5): aiming at the space-time characteristic index, establishing a ConvLSTM network space-time characteristic extraction model;
(6): aiming at the time characteristic index, establishing an SRU network time characteristic extraction model;
(7): establishing an index characteristic full-connection network model;
(8): outputting a predicted sector flight time;
(9): and calculating the predicted delay time and prediction accuracy of the sector.
Further, the specific process of the step (5) is as follows:
(5.1) generating ConvLSTM network input set
Counting the number of the fan entering and fan leaving waypoints as m and n, making a picture with the size of m multiplied by n, wherein the numerical value of each pixel point is the flight flow of a corresponding flight path between the fan entering and fan leaving points, generating an original picture of the flight flow at a corresponding time interval according to the time sequence data obtained in the step (4), and generating the original picture of the flight time at the corresponding time interval by taking the flight time of the flight path between the fan entering and fan leaving points as the pixel point;
(5.2) initializing ConvLSTM network parameters
(5.2.1) initializing input layer parameters
When constructing the input layer of the convLSTM network, setting the initial sample number of each batch of training and the initial setting value of the time step, preferably, setting the initial sample number of each batch of training to be 48; setting the initial value of the time step length as 6 time intervals;
(5.2.2) initializing network layer parameters
When a network layer of the ConvLSTM network is constructed, setting initial setting values of convolution dimensionality, input dimensionality, output depth and convolution kernel size, and preferably setting the initial setting value of the convolution dimensionality to be 2; setting an initial setting value of an input dimension to be [ m, n,1 ]; setting an initial setting value of the output depth to be 6; setting the initial size of a convolution kernel as [3,3 ]; the activation function initially selects Relu; initializing weights and biases of a network layer;
(5.3) setting input sample and output characteristics of ConvLSTM network
(5.3.1) construction of input samples
Respectively dividing the original pictures of the flight flow and the flight time of the flight route obtained in the step (5.1) into a plurality of time sequences according to the initial time step length set in the step (5.2.1);
(5.3.2) construction of output features
The network output characteristic is used for acquiring a characteristic value of the last time step;
(5.3.3) data normalization
And (4) performing min-max normalization on the input sample to generate a dimensionless training data set.
Further, the specific process of step (6) is as follows:
(6.1) initializing SRU network parameters;
(6.1.1) initializing input layer parameters
When an input layer of the SRU network is constructed, giving the initial setting values of the sample number and the time step length of each batch of training; preferably, the initial sample number of each batch of training is set to be 48; setting the initial value of the time step length as 6 time intervals; (6.1.2) initializing network layer parameters
When constructing the network layer of the SRU network, giving the initial setting values of the number of the network layers and the number of neurons in each layer; preferably, setting the initial setting value of the network layer number to be 2; setting the initial value of each layer of neurons as 50; the activation function initially selects Relu, and initializes the weight and bias of the network layer;
(6.2) setting input sample and output characteristics of SRU network
(6.2.1) construction of input samples
Respectively dividing the flight flow and meteorological data sequence of the fan entering and leaving waypoints obtained in the step (4) into a plurality of time sequences according to the initial time step length set in the step (6.1.1);
(6.2.2) construction of output features
The network output characteristic is used for acquiring a characteristic value of the last time step;
(6.2.3) data normalization
And (4) performing min-max normalization on the input sample to generate a dimensionless training data set.
Further, the specific process of step (7) is as follows:
(7.1) initializing index feature fully-connected network parameters
(7.1.1) initializing weights and biases of the input layers;
(7.1.2) initially selecting Relu as an output layer activation function;
(7.2) setting input samples and output samples of the index feature fully-connected network
(7.2.1) construction of input samples
Converting the space-time characteristic value output in the step (5.3.2) into a one-dimensional vector, and using the one-dimensional vector and the time characteristic extracted in the step (6.2.2) as an input sample of the fully-connected network;
(7.2.2) construction of output samples
Taking the data of the last time step of the flight time sequence obtained in the step (5.3.1) as an output sample of the index feature fully-connected network;
(7.2.3) data normalization
And (4) performing min-max normalization on the output samples to generate a dimensionless training data set.
Further, the specific process of step (8) is as follows:
(8.1) setting training set and test set
Randomly extracting the data set x k (0< k <1) obtained in the step (4) as a training set, and using the rest as a testing set, wherein k is preferably 0.8;
(8.2) loss function setting
Selecting a square error loss function in a sector airline time of flight prediction neural network;
(8.3) optimizer settings
The optimizer selects Adam, Adagrad, RMSprop, NAG and SGD respectively to perform a comparison experiment, and preferably selects Adam;
(8.4) data de-normalization
And (4) carrying out the inverse operation of min-max normalization on the output value of the test set, wherein the obtained numerical value is the predicted flight time.
Further, the specific process of step (9) is as follows:
(9.1) calculating a sector flight path time-of-flight reference value
Sequencing the flight time data of each air route of the sector obtained in the step (2) from small to large, and selecting a flight time reference value, wherein the preferred flight time reference value is a numerical value corresponding to 20% quantile points;
(9.2) calculating the expected delay time of the sector
Calculating the difference between the predicted flight time obtained in the step (8.4) and the flight time reference value obtained in the step (9.1) to obtain the predicted delay time of the sector;
(9.3) calculating the prediction accuracy
And selecting the average error ratio (MAPE) as an evaluation index, and calculating the prediction precision of the sector delay time.
Further, the method also comprises a step (10), and the specific process is as follows:
(10.1) calculating prediction accuracy of LSTM and GRU networks
Expanding the two-dimensional data generated in the step (5.1) to form one-dimensional data, combining the one-dimensional data with the data in the step (6), respectively inputting the data into an LSTM network and a GRU network for prediction, and calculating prediction precision;
(10.2) prediction accuracy comparison
And (4) comparing the prediction precision of the sector delay time calculated in the step (9.3) with the prediction precision of the common deep learning method (LSTM, GRU) in the step (10.1).
Has the advantages that: the ConvLSTM-SRU-based sector delay prediction method provided by the invention can improve the prediction precision of sector delay time, provide prospective information for air traffic control personnel, help the control personnel to take control measures in advance, reduce the probability of increased sector congestion risks and further improve the operation efficiency of an airport. Compared with the traditional method for deducing the delay time based on the flow, the method considers the factors of the flow and the navigation network structure of the sector, has better adaptive performance, and has practical engineering application value in the aspect of predicting the delay time of the sector.
Drawings
FIG. 1 is a flow chart of a ConvLSTM-SRU-based sector stall prediction method of the present invention.
Detailed Description
For a further understanding of the present invention, reference will now be made in detail to the embodiments illustrated in the drawings.
The method for predicting the sector delay based on the ConvLSTM-SRU comprises the following steps:
(1) reading data: reading historical flight ADS-B data of a sector, counting to obtain flight flow data of each waypoint of an entering sector and an leaving sector, flight flow data and flight time data of each airline (from the point of entering the sector to the point of leaving the sector) in the sector, reading historical meteorological data of the sector in a corresponding time range, and counting to obtain partial experimental data when a time interval is 10min in a table 1 and a table 2;
TABLE 1
TABLE 2
(2) Processing data: arranging the data sets read in the step (1) according to time tags, filling missing data on a time axis by using the mean values of the time periods before and after, and replacing abnormal data on the time axis by using an interpolation method;
(3) dividing index characteristics: the method comprises the following steps of performing type division on indexes according to index attributes, and dividing the two indexes into space-time characteristic indexes as the flight flow and the flight time of each route (from a sector entering point to a sector leaving point) in a sector change along with time and space (a sector route network); the flight flow and the meteorological conditions of each waypoint entering and leaving the sector only change along with time in the same sector and are irrelevant to the space, so the two indexes are divided into time characteristic indexes;
(4) generating time-series data of each index: on the basis of the step (3), selecting different time intervals (10min, 20min, 30min and 60min), and carrying out statistics on data of each index to generate sequence data of corresponding time intervals;
(5): establishing a ConvLSTM network space-time feature extraction model aiming at space-time characteristic indexes
(5.1) generating ConvLSTM network input set
Counting 5 and 4 fan-entering and fan-leaving waypoints to produce pictures with the size of 5 multiplied by 4, wherein the numerical value of each pixel point is the flight flow of a flight path between the corresponding fan-entering and fan-leaving points, generating the original pictures of the flight flow at the corresponding time interval according to the time sequence data obtained in the step (4), and generating the original pictures of the flight time at the corresponding time interval by taking the flight time of the flight path between the fan-entering and fan-leaving points as the pixel point;
(5.2) initializing ConvLSTM network parameters
(5.2.1) initializing input layer parameters
Setting the initial sample number of each batch of training to be 48 when constructing the input layer of the ConvLSTM network; setting the initial value of the time step length as 6 time intervals;
(5.2.2) initializing network layer parameters
When a network layer of the ConvLSTM network is constructed, setting an initial setting value of convolution dimension to be 2; setting an initial setting value of an input dimension to be [5,4,1 ]; setting an initial setting value of the output depth to be 6; setting the initial size of a convolution kernel as [3,3 ]; the activation function initially selects Relu; initializing weights and biases of a network layer;
(5.3) setting input sample and output characteristics of ConvLSTM network
(5.3.1) construction of input samples
Respectively dividing the original pictures of the flight flow and the flight time of the flight route obtained in the step (5.1) into a plurality of time sequences according to the initial time step length set in the step (5.2.1);
(5.3.2) construction of output features
The network output characteristic is used for acquiring a characteristic value of the last time step;
(5.3.3) data normalization
Performing min-max normalization on the input sample to generate a dimensionless training data set;
(6): aiming at the time characteristic index, establishing an SRU network time characteristic extraction model
(6.1) initializing SRU network parameters;
(6.1.1) initializing input layer parameters
When an input layer of the SRU network is constructed, setting the initial sample number of each batch of training to be 48; setting the initial value of the time step length as 6 time intervals; (ii) a
(6.1.2) initializing network layer parameters
When a network layer of the SRU network is constructed, setting an initial setting value of the number of the network layers to be 2; setting the initial value of each layer of neurons as 50; the activation function initially selects Relu, and initializes the weight and bias of the network layer;
(6.2) setting input sample and output characteristics of SRU network
(6.2.1) construction of input samples
Respectively dividing the flight flow and meteorological data sequence of the fan entering and leaving waypoints obtained in the step (4) into a plurality of time sequences according to the initial time step length set in the step (6.1.1);
(6.2.2) construction of output features
The network output characteristic is used for acquiring a characteristic value of the last time step;
(6.2.3) data normalization
Performing min-max normalization on the input sample to generate a dimensionless training data set;
(7): establishing index characteristic full-connection network model
(7.1) initializing index feature fully-connected network parameters
(7.1.1) initializing weights and biases of the input layers;
(7.1.2) initially selecting Relu as an output layer activation function;
(7.2) setting input samples and output samples of the index feature fully-connected network
(7.2.1) construction of input samples
Converting the space-time characteristic value output in the step (5.3.2) into a one-dimensional vector, and using the one-dimensional vector and the time characteristic extracted in the step (6.2.2) as an input sample of the fully-connected network;
(7.2.2) construction of output samples
Taking the data of the last time step of the flight time sequence obtained in the step (5.3.1) as an output sample of the index feature fully-connected network;
(7.2.3) data normalization
Performing min-max normalization on the output samples to generate a dimensionless training data set;
(8): output predicted sector time of flight
(8.1) setting training set and test set
Randomly extracting 80% of the data set obtained in the step (4) as a training set, and taking the rest 20% as a test sample;
(8.2) loss function setting
Selecting a square error loss function in a sector airline time of flight prediction neural network;
(8.3) optimizer settings
The optimizer selects Adam;
(8.4) data de-normalization
Performing inverse operation of min-max normalization on the output value of the test set, wherein the obtained numerical value is the predicted flight time;
(9): calculating the predicted delay time and prediction accuracy of a sector
(9.1) calculating a sector flight path time-of-flight reference value
Sequencing the flight time data of each air route of the sector obtained in the step (2) from small to large, and selecting a numerical value corresponding to a 20% quantile point as a flight time reference value;
(9.2) calculating the expected delay time of the sector
Calculating the difference between the predicted flight time obtained in the step (8.4) and the flight time reference value obtained in the step (9.1) to obtain the predicted delay time of the sector;
(9.3) calculating the prediction accuracy
Selecting an average error ratio (MAPE) as an evaluation index, and calculating the prediction precision of the sector delay time;
(10): comparing the prediction precision;
(10.1) calculating prediction accuracy of LSTM and GRU networks
Expanding the two-dimensional data generated in the step (5.1) to form one-dimensional data, combining the one-dimensional data with the data in the step (6), respectively inputting the data into an LSTM network and a GRU network for prediction, and calculating prediction precision;
(10.2) prediction accuracy comparison
Comparing the sector delay prediction result based on ConvLSTM-SRU calculated in the step (9.3) with the prediction accuracy of the common deep learning methods (LSTM, GRU) calculated in the step (10.1), as shown in Table 3, the method provided by the invention can improve the sector delay time prediction accuracy.
TABLE 3
Claims (6)
1. A ConvLSTM-SRU-based sector delay prediction method is characterized by comprising the following steps:
(1) reading data: reading historical flight ADS-B data of the sector, counting to obtain flight flow data of each waypoint entering the sector and leaving the sector, flight flow data and flight time data of each airline in the sector, and reading historical meteorological data of the sector in a corresponding time range;
(2) processing data: arranging the data sets read in the step (1) according to time tags, filling missing data on a time axis by using the mean values of the time periods before and after, and replacing abnormal data on the time axis by using an interpolation method;
(3) dividing index characteristics: the method comprises the following steps of performing type division on indexes according to index attributes, and dividing the flight flow and the flight time of each air route in a sector into space-time characteristic indexes as the flight flow and the flight time of each air route in the sector change along with time and space; the flight flow and the meteorological conditions of each waypoint of the entering and leaving sector only change along with time in the same sector and are irrelevant to the space, so the flight flow and the meteorological conditions of each waypoint of the entering and leaving sector are divided into time characteristic indexes;
(4) generating time-series data of each index: on the basis of the step (3), selecting different time intervals, counting the data of each index, and generating sequence data of the corresponding time intervals;
(5): aiming at the space-time characteristic index, establishing a ConvLSTM network space-time characteristic extraction model;
(6): aiming at the time characteristic index, establishing an SRU network time characteristic extraction model;
(7): establishing an index characteristic full-connection network model;
(8): outputting a predicted sector flight time;
(9): and calculating the predicted delay time and prediction accuracy of the sector.
2. The ConvLSTM-SRU-based sector delay prediction method of claim 1, wherein the specific process of step (5) is as follows:
(5.1) generating ConvLSTM network input set
Counting the number of fan entering and fan leaving waypoints as m and n, and making a picture with the size of m multiplied by n, wherein the numerical value of each pixel point is the flight flow of a corresponding flight path between the fan entering and fan leaving points, the time sequence data obtained in the step (4) is used as the pixel point to generate an original picture of the flight flow at a corresponding time interval, and similarly, the flight time of the flight path between the fan entering and fan leaving points is used as the pixel point to generate the original picture of the flight time at the corresponding time interval;
(5.2) initializing ConvLSTM network parameters
(5.2.1) initializing input layer parameters
When an input layer of the ConvLSTM network is constructed, giving initial setting values of the number of samples and the time step of each batch of training;
(5.2.2) initializing network layer parameters
When a network layer of the ConvLSTM network is constructed, setting initial setting values of a convolution dimension, an input dimension, an output depth and a convolution kernel size; the activation function initially selects Relu, and initializes the weight and bias of the network layer;
(5.3) setting input sample and output characteristics of ConvLSTM network
(5.3.1) construction of input samples
Respectively dividing the original pictures of the flight flow and the flight time of the flight route obtained in the step (5.1) into a plurality of time sequences according to the initial time step length set in the step (5.2.1);
(5.3.2) construction of output features
The network output characteristic is used for acquiring a characteristic value of the last time step;
(5.3.3) data normalization
And (4) performing min-max normalization on the input sample to generate a dimensionless training data set.
3. The ConvLSTM-SRU-based sector delay prediction method of claim 1, wherein the specific process of step (6) is as follows:
(6.1) initializing SRU network parameters;
(6.1.1) initializing input layer parameters
When an input layer of the SRU network is constructed, giving the initial setting values of the sample number and the time step length of each batch of training; (6.1.2)
Initializing network layer parameters
When constructing the network layer of the SRU network, giving the initial setting values of the number of the network layers and the number of neurons in each layer; the activation function initially selects Relu, and initializes the weight and bias of the network layer;
(6.2) setting input sample and output characteristics of SRU network
(6.2.1) construction of input samples
Respectively dividing the flight flow and meteorological data sequence of the fan entering and leaving waypoints obtained in the step (4) into a plurality of time sequences according to the initial time step length set in the step (6.1.1);
(6.2.2) construction of output features
The network output characteristic is used for acquiring a characteristic value of the last time step;
(6.2.3) data normalization
And (4) performing min-max normalization on the input sample to generate a dimensionless training data set.
4. The ConvLSTM-SRU-based sector delay prediction method of claim 1, wherein the specific process of step (7) is as follows:
(7.1) initializing index feature fully-connected network parameters
(7.1.1) initializing weights and biases of the input layers;
(7.1.2) initially selecting Relu as an output layer activation function;
(7.2) setting input samples and output samples of the index feature fully-connected network
(7.2.1) construction of input samples
Converting the space-time characteristic value output in the step (5.3.2) into a one-dimensional vector, and using the one-dimensional vector and the time characteristic extracted in the step (6.2.2) as an input sample of the fully-connected network;
(7.2.2) construction of output samples
Taking the data of the last time step of the flight time sequence obtained in the step (5.3.1) as an output sample of the index feature fully-connected network;
(7.2.3) data normalization
And (4) performing min-max normalization on the output samples to generate a dimensionless training data set.
5. The ConvLSTM-SRU-based sector delay prediction method of claim 1, wherein the specific process of step (8) is as follows:
(8.1) setting training set and test set
Randomly extracting a data set xk obtained in the step (4) as a training set, wherein k is 0< k <1, and the rest is used as a test set;
(8.2) loss function setting
Selecting a square error loss function in a sector airline time of flight prediction neural network;
(8.3) optimizer settings
The optimizer selects Adam;
(8.4) data de-normalization
And (4) carrying out the inverse operation of min-max normalization on the output value of the test set, wherein the obtained numerical value is the predicted flight time.
6. The ConvLSTM-SRU-based sector delay prediction method of claim 1, wherein the specific process of step (9) is as follows:
(9.1) calculating a sector flight path time-of-flight reference value
Sequencing the flight time data of each air route of the sector obtained in the step (2) from small to large, and selecting a flight time reference value;
(9.2) calculating the expected delay time of the sector
Calculating the difference between the predicted flight time obtained in the step (8.4) and the flight time reference value obtained in the step (9.1) to obtain the predicted delay time of the sector;
(9.3) calculating the prediction accuracy
And selecting MAPE as an evaluation index, and calculating the prediction precision of the sector delay time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010277615.4A CN111523641B (en) | 2020-04-10 | 2020-04-10 | Sector delay prediction method based on ConvLSTM-SRU |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010277615.4A CN111523641B (en) | 2020-04-10 | 2020-04-10 | Sector delay prediction method based on ConvLSTM-SRU |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111523641A true CN111523641A (en) | 2020-08-11 |
| CN111523641B CN111523641B (en) | 2023-05-30 |
Family
ID=71902708
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010277615.4A Active CN111523641B (en) | 2020-04-10 | 2020-04-10 | Sector delay prediction method based on ConvLSTM-SRU |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111523641B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105225007A (en) * | 2015-09-30 | 2016-01-06 | 中国民用航空总局第二研究所 | A kind of sector runnability method for comprehensive detection based on GABP neural network and system |
| CN105679102A (en) * | 2016-03-03 | 2016-06-15 | 南京航空航天大学 | National flight flow space-time distribution prediction deduction system and method |
| CN106023655A (en) * | 2016-06-30 | 2016-10-12 | 南京航空航天大学 | Sector air traffic congestion state monitoring method |
| CN109658741A (en) * | 2018-12-12 | 2019-04-19 | 中国船舶重工集团公司第七0九研究所 | A kind of sector short term traffic forecasting method and system |
-
2020
- 2020-04-10 CN CN202010277615.4A patent/CN111523641B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105225007A (en) * | 2015-09-30 | 2016-01-06 | 中国民用航空总局第二研究所 | A kind of sector runnability method for comprehensive detection based on GABP neural network and system |
| CN105679102A (en) * | 2016-03-03 | 2016-06-15 | 南京航空航天大学 | National flight flow space-time distribution prediction deduction system and method |
| CN106023655A (en) * | 2016-06-30 | 2016-10-12 | 南京航空航天大学 | Sector air traffic congestion state monitoring method |
| CN109658741A (en) * | 2018-12-12 | 2019-04-19 | 中国船舶重工集团公司第七0九研究所 | A kind of sector short term traffic forecasting method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111523641B (en) | 2023-05-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113657465B (en) | Pre-training model generation method and device, electronic equipment and storage medium | |
| CN114037844A (en) | Global rank perception neural network model compression method based on filter characteristic diagram | |
| CN110164129B (en) | Single-intersection multi-lane traffic flow prediction method based on GERNN | |
| CN112489497B (en) | Airspace operation complexity evaluation method based on deep convolutional neural network | |
| CN109492748B (en) | Method for establishing medium-and-long-term load prediction model of power system based on convolutional neural network | |
| CN111626366B (en) | Operation characteristic-based area sector scene similarity identification method | |
| CN110442143B (en) | Unmanned aerial vehicle situation data clustering method based on combined multi-target pigeon swarm optimization | |
| CN113341919B (en) | Computing system fault prediction method based on time sequence data length optimization | |
| CN112994701A (en) | Data compression method and device, electronic equipment and computer readable medium | |
| CN114925238B (en) | Federal learning-based video clip retrieval method and system | |
| CN116468186A (en) | Flight delay time prediction method, electronic equipment and storage medium | |
| CN116244647A (en) | Unmanned aerial vehicle cluster running state estimation method | |
| CN116109195A (en) | Performance evaluation method and system based on graph convolution neural network | |
| CN114897085A (en) | Clustering method based on closed subgraph link prediction and computer equipment | |
| CN112860685A (en) | Automatic recommendation of analysis of data sets | |
| CN116151479B (en) | Flight delay prediction method and prediction system | |
| CN111523641A (en) | ConvLSTM-SRU-based sector delay prediction method | |
| CN115759470A (en) | Flight overall process fuel consumption prediction method based on machine learning | |
| CN115600744A (en) | Method for predicting population quantity of shared space-time attention convolutional network based on mobile phone data | |
| CN113743453A (en) | Population quantity prediction method based on random forest | |
| CN112418730A (en) | Construction method of response index estimation model of transportation system | |
| CN117422320B (en) | Method for extracting influence factors of weather on flight toughness operation | |
| CN111652102A (en) | Power transmission channel target object identification method and system | |
| CN110941767A (en) | Network community detection countermeasure enhancement method based on multi-similarity integration | |
| CN113222229B (en) | Non-cooperative unmanned aerial vehicle track prediction method based on machine learning |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |