CN114037159A - Short-time passenger flow prediction method and system based on multi-source data fusion input - Google Patents
Short-time passenger flow prediction method and system based on multi-source data fusion input Download PDFInfo
- Publication number
- CN114037159A CN114037159A CN202111322660.8A CN202111322660A CN114037159A CN 114037159 A CN114037159 A CN 114037159A CN 202111322660 A CN202111322660 A CN 202111322660A CN 114037159 A CN114037159 A CN 114037159A
- Authority
- CN
- China
- Prior art keywords
- data
- passenger flow
- flow prediction
- source data
- sample 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.)
- Pending
Links
- 230000004927 fusion Effects 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000000605 extraction Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 5
- 239000013598 vector Substances 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
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 230000015654 memory Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000001502 supplementing effect Effects 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000006403 short-term memory Effects 0.000 claims 1
- 238000007500 overflow downdraw method Methods 0.000 abstract description 2
- 230000008859 change Effects 0.000 description 10
- 238000012546 transfer Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000001507 sample dispersion Methods 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000002759 z-score normalization Methods 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2457—Query processing with adaptation to user needs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- 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
- G06N3/048—Activation functions
-
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Human Resources & Organizations (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Primary Health Care (AREA)
- Bioinformatics & Computational Biology (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Fuzzy Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a short-time passenger flow prediction method based on multi-source data fusion input, which comprises the following steps of: acquiring sample data; preprocessing sample data; extracting multi-source data characteristics from sample data; the dimension of the obtained external factor characteristic and the dimension of the video characteristic are adjusted through two full connection layers, so that the dimension of each characteristic is unified; performing multi-source data characteristic fusion on the sample data, and establishing a passenger flow prediction model; and predicting the passenger flow data according to the passenger flow prediction model, and automatically adjusting the working states of various facilities according to the prediction condition. The invention discloses a short-time passenger flow prediction system based on multi-source data fusion input. In the invention, uniform data characteristics are extracted from the complicated passenger flow influence factors by a multi-source data fusion method, so that the accuracy of short-time passenger flow prediction is improved, the accurate passenger flow prediction and alarm of different specific scenes are realized, the working states of various facilities are automatically adjusted according to the prediction conditions, and the safe operation is ensured.
Description
Technical Field
The invention relates to the technical field of rail transit, in particular to a short-time passenger flow prediction method and system based on multi-source data fusion input.
Background
The rail transit passenger flow has obvious spatiotemporal and periodicity, is easily influenced by factors such as weather, emergencies, holidays, large activities and the like, and in order to accurately predict the rail transit short-term passenger flow, it is necessary to analyze and research multi-source data influencing the passenger flow. The multi-source data has very strong space-time attribute and is often accompanied with the problems of multi-source and isomerism of the data, wherein the source channels of the multi-source data are various, and the isomerism data comprises structured data and unstructured data. According to the type of the data structure, the multi-source data fusion is divided into isomorphic multi-source data fusion and heterogeneous multi-source data fusion. The isomorphic multi-source data refers to that the sources of the multi-source data belong to the same field category, and the heterogeneous multi-source data refers to that the data come from different field categories in the physical world, and can accurately describe targets in respective fields, but mutually overlapped parts exist when a certain potential target is represented. For example, geographic locations, weather, emergencies and the like belong to different fields, and describe different targets, so that the data cannot be directly fused, and data features in the data cannot be extracted and then fused.
An immature place exists in the current multi-source data fusion technology, the data processing and preprocessing, correlation analysis or direct fusion often stays, and the dynamic change process of each factor influencing the rail transit on the space and time is ignored.
Disclosure of Invention
In order to solve the technical problems in the background art, the invention provides a short-time passenger flow prediction method and system based on multi-source data fusion input, and aims to solve the problems of poor multi-source data fusion efficiency and low rail transit passenger flow prediction accuracy.
The invention provides a short-time passenger flow prediction method based on multi-source data fusion input, which comprises the following steps of:
101, acquiring sample data comprising space-time characteristic data, external factor characteristic data and recent video data;
102, preprocessing sample data;
103, extracting multi-source data characteristics from the sample data, including the extraction of time-space characteristics, external factor characteristics and video characteristics;
104, adjusting the obtained dimensions of the external factor characteristic and the video characteristic through two full connection layers to enable the dimensions of all the characteristics to be uniform;
105, performing multi-source data feature fusion on the sample data, and establishing a passenger flow prediction model;
and 106, predicting the passenger flow data according to the passenger flow prediction model, and automatically adjusting the working states of various facilities according to the prediction condition.
The invention provides a short-time passenger flow prediction system based on multi-source data fusion input, which comprises:
the data acquisition module is used for acquiring sample data, including space-time characteristic data, external factor characteristic data and recent video data;
the preprocessing module is connected with the data acquisition module and is used for preprocessing the sample data;
the characteristic extraction module is connected with the preprocessing module and used for extracting multi-source data characteristics from the sample data, including the extraction of time-space characteristics, external factor characteristics and video characteristics;
the dimension adjusting module is connected with the feature extracting module and used for adjusting the obtained dimensions of the external factor features and the video features through two full connecting layers to ensure that the dimensions of all the features are uniform;
the multi-source data feature fusion module is connected with the dimension adjusting module and used for carrying out multi-source data feature fusion on the sample data and establishing a passenger flow prediction model for obtaining passenger flow prediction;
and the passenger flow prediction module is connected with the multi-source data characteristic fusion module and used for predicting passenger flow data by the passenger flow prediction module and automatically adjusting the working states of various facilities according to the prediction condition.
In the invention, uniform data characteristics are extracted from the complicated passenger flow influence factors by a multi-source data fusion method, so that the accuracy of short-time passenger flow prediction is improved, the accurate passenger flow prediction and alarm of different specific scenes are realized, the working states of various facilities are automatically adjusted according to the prediction conditions, and the safe operation is ensured.
Drawings
Fig. 1 is a flowchart of a short-term passenger flow prediction method based on multi-source data fusion input according to an embodiment of the present invention;
fig. 2 is a structural diagram of a short-term passenger flow prediction system based on multi-source data fusion input according to an embodiment of the present invention.
Detailed Description
The invention provides a short-time passenger flow prediction method based on multi-source data fusion input, as shown in figure 1, comprising the following steps:
the space-time characteristic data mainly comprises: the recent (such as five years) real-time passenger flow data of the network, line and station, such as the station entering and exiting amount, the passenger transportation amount, the transfer amount, the OD amount, the transfer imbalance coefficient and the like, and the passenger flow clearing data.
The external factor characteristic data mainly comprises: the recent holiday data comprise New year's day, Qingming day, Wuyi day, Dragon's day, mid-autumn day, eleven and spring festival, etc.; recent actual weather data and weather forecast data in a future forecast period, wherein the weather forecast data comprises temperature, weather, wind direction, rainfall, humidity, air quality and the like; relevant data of recent new line opening; recent operation emergency data, large-scale activity data in a prediction period, instant operation emergency data and the like; recent car restriction data, public transportation and subway data, mobile phone signaling data and the like;
step 1021, selecting input sample data. The method specifically comprises the following steps: checking whether the net structure changes in the time range from the time beginning of the selected sample data to the time ending to be predicted, if so, then: firstly, the sample data volume after the change of the net structure reaches the requirement of passenger flow prediction model input, and abandoning the sample data before the change of the net structure; secondly, the sample data size is sufficient after the change of the net structure or the change of the net structure occurs after the sample data time range and before the prediction time range, then the passenger flow change coefficient (mu) is calculated according to the passenger flow change caused by the change of the net structure1) And the coefficient of variation (mu) of passenger flow and passenger flow caused by the change of the wire network structure are used together1) The product of (a) and (b) is used as input sample data;
step 1022, checking whether the weather condition of the date corresponding to the sample data is the same as or similar to the weather of the date to be predicted; if the weather of the date corresponding to the historical statutory holiday is different from the weather of the forecast day, the same weather is used for the first timeThe same-date passenger flow data and passenger flow variation coefficient (mu) which are the same or similar to the weather of the historical legal holidays under the network structure2) Supplementing the product of (A); second, using the average value of friday/saturday/sunday/monday of the same or similar weather history under the same wire net structure and the passenger flow variation coefficient (mu)3) Supplementing the product of (A);
step 1023, checking whether the sample data has missing value, abnormal value, extreme value, discrete value, etc., if yes, using the average value or the average value of the passenger flow data and the passenger flow variation coefficient (mu) of the same week and the same or similar weather date in the historical legal holiday2) The product of (a) and (b) is supplemented.
Another data preprocessing method in the embodiment of the present invention is:
a. checking the integrity of input sample data, and checking whether missing values, abnormal values, extreme values, discrete values and the like exist in the input data; if so, performing interpolation by using the mean value;
b. carrying out characteristic One-hot Encoding (One-hot Encoding) processing on external factor data, wherein the One-hot Encoding can process non-numerical attributes such as holidays, weather and the like; features are expanded to a certain extent; the coded attributes are sparse, with a large number of zero components;
c. data standardization and regularization, wherein the standardization is to scale the data attribute to a specified range so that sample data has zero mean and unit variance; data were normalized to min-max and z-score.
min-max normalization:
wherein x is*The value is the value after data processing, x is the original data, max is the maximum value of the data, and min is the minimum value of the data; min-max normalization is the mapping of data to [0,1]In the meantime.
z-score normalization:
wherein x is*Is the value after data processing, x is the original data,
the mean value of the original data, and sigma is the standard deviation of the original data; the mean of the data normalized by z-score was 0 and the standard deviation was 1.
characteristics of space and time
If the transfer amount is predicted, using historical transfer amount data; if the passenger volume is predicted, the historical passenger volume is used. Abstracting historical passenger flow data into vectors
X＝[xt-1，xt-2，xt-3，xt-4，xt-5，xt-6，xt-7，xt-14，xt-21，xt-28，x]
The method for extracting the time sequence characteristics by using the long-short term memory network LSTM comprises the following steps:
1) forgetting stage
The forgetting gate processes historical output and current input data through the sigmoid activation function to determine the cell state C at the previous momentt-1And keeping the information of the current time.
ft＝σ(Wf·[ht-1，X]+bf)
In the formula (f)tIs the state value of the forgetting gate, sigma is the sigmoid activation function, ht-1Is the output at the previous time, X is the input data, WfAnd bfRespectively representing the weight matrix and the bias terms.
2) Memory stage
The input gate determines the newly added input information of the network at the current moment through the sigmoid activation function, and in addition, a new candidate vector is created based on the historical output and the current input data through the tanh activation function
it＝σ(Wi·[ht-1，X]+bi)
The contents of the cell state are then updated by both the forgetting gate and the input gate.
In the formula itIs the status value of the input gate, Ct-1Is a historical cell state, CtIs the current cell state, WiAnd WcIs a weight matrix, biAnd bcIs the bias term.
3) Output stage
And finally, outputting the output information determined by the sigmoid activation function, and calculating the output value of the LSTM at the current moment based on the updated cell state.
Ot＝σ(W0·[ht-1，X]+b0)
ht＝ot*tanh(Ct)
In the formula, OtIs the state value of the output gate, W0And b0Respectively representing the weight matrix and the bias term, htIs the updated hidden layer state. Represents the hadamard product operation.
External factor characteristics
For data such as holidays, weather, emergencies and the like, onehot coding is firstly carried out, and then external features are extracted by using a full-connection network;
thirdly, extracting the video features by using a convolutional neural network CNN.
and the primary features extracted from the external data further extract the external features by using two full-connection layers, wherein the first layer is an embedding layer of each type of data, and the second layer carries out dimension increasing on the embedded vector, so that the dimensions of the multi-source heterogeneous data are kept the same.
And 105, performing multi-source data feature fusion on the sample data, and establishing passenger flow prediction to obtain a passenger flow prediction model. Output H extracted from LSTMtFusing with external features to obtain the input of joint prediction with the prediction result of
The goal of model training is to minimize the error between true passenger flow and predicted passenger flow. A loss function of
Wherein the content of the first and second substances,
for minimizing error, λLregIs L2And the regularization term is helpful for solving the overfitting problem of the model.
And 106, evaluating the passenger flow prediction model. In order to evaluate the prediction effect of the model, the average absolute error (MAE), the Root Mean Square Error (RMSE) and the average absolute percentage error (MAPE) are used as the performance evaluation indexes of the model, and the calculation formula is as follows:
wherein i is the ith data,
as a predicted value, YiN is the number of data for the original data.
The Mean Absolute Error (MAE) is the average of the absolute values of the deviations of all individual observations from the arithmetic mean. The average absolute error can avoid the problem of mutual offset of errors, so that the size of the actual prediction error can be accurately reflected. The advantage is a better reflection of the actual situation of the error.
The Root Mean Square Error (RMSE) represents the standard deviation of the samples of the difference between the predicted and observed values, called the residual. The root mean square error is to illustrate the degree of sample dispersion. When fitting non-linearly, the smaller the RMSE, the better. The root mean square error is the deviation of the observed value from its analog value, not the deviation of the observed value from its average value. While RMSE penalizes higher discrepancies more than MAE. The average absolute percentage error is often used as a statistical measure of prediction accuracy, such as time series predictions. The smaller the value of MAPE, the better the accuracy of the prediction model.
And step 107, predicting the passenger flow data according to the passenger flow prediction model, and automatically adjusting the working states of various facilities according to the prediction condition.
The embodiment provides a short-time passenger flow prediction system based on multi-source data fusion input, as shown in fig. 2, including:
the data acquisition module is used for acquiring sample data, including space-time characteristic data, external factor characteristic data and recent video data;
the preprocessing module is connected with the data acquisition module and is used for preprocessing the sample data;
the characteristic extraction module is connected with the preprocessing module and used for extracting multi-source data characteristics from the sample data, including the extraction of time-space characteristics, external factor characteristics and video characteristics;
the dimension adjusting module is connected with the feature extracting module and used for adjusting the obtained dimensions of the external factor features and the video features through two full connecting layers to ensure that the dimensions of all the features are uniform;
the multi-source data feature fusion module is connected with the dimension adjusting module and used for carrying out multi-source data feature fusion on the sample data and establishing a passenger flow prediction model for obtaining passenger flow prediction;
and the passenger flow prediction module is connected with the multi-source data characteristic fusion module and used for predicting passenger flow data by the passenger flow prediction module and automatically adjusting the working states of various facilities according to the prediction condition. And evaluating the passenger flow prediction model, wherein the average absolute error (MAE), the Root Mean Square Error (RMSE) and the average absolute percentage error (MAPE) are used as the performance evaluation indexes of the model.
According to the invention, by collecting and analyzing various factor data influencing passenger flow, and fully considering influence factors and influence weights influencing passenger flow change, data characteristics are further mined and extracted, and the data characteristics are input into an advanced machine learning algorithm, so that the passenger flow prediction accuracy is improved, the working states of various facilities are automatically adjusted according to the prediction condition, and safe operation is ensured. For example, when the passenger flow is predicted to exceed the safety threshold, the standby gate and the passage are opened, the automatic guidance indication is started, and passengers are guided to shunt in and out of the gate or the platform entrance in the modes of voice, graphic representation and the like.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (7)
1. A short-time passenger flow prediction method based on multi-source data fusion input is characterized by comprising the following steps:
101, acquiring sample data comprising space-time characteristic data, external factor characteristic data and recent video data;
102, preprocessing sample data;
103, extracting multi-source data characteristics from the sample data, including the extraction of time-space characteristics, external factor characteristics and video characteristics;
104, adjusting the obtained dimensions of the external factor characteristic and the video characteristic through two full connection layers to enable the dimensions of all the characteristics to be uniform;
105, performing multi-source data feature fusion on the sample data, and establishing a passenger flow prediction model;
and 106, predicting the passenger flow data according to the passenger flow prediction model, and automatically adjusting the working states of various facilities according to the prediction condition.
2. The multi-source data fusion input-based short-time passenger flow prediction method according to claim 1, wherein sample data is preprocessed, specifically comprising:
selecting input sample data;
checking whether the weather condition of the date corresponding to the sample data is the same as or similar to the weather of the date to be predicted or not; if there is a big difference between the weather of the date corresponding to the historical statutory holiday and the weather of the forecast day, then firstly, the same-date passenger flow data and the passenger flow variation coefficient (mu) which are the same or similar to the weather of the historical statutory holiday under the same network structure are used2) Supplementing the product of (A); second, using the average value of friday/saturday/sunday/monday of the same or similar weather history under the same wire net structure and the passenger flow variation coefficient (mu)3) Supplementing the product of (A);
checking whether the sample data has missing value, abnormal value, extreme value, discrete value, etc., if yes, using the mean value or mean value of the passenger flow data and the passenger flow variation coefficient (mu) of the same week and the same or similar weather date in the historical legal holiday2) The product of (a) and (b) is supplemented.
3. The multi-source data fusion input-based short-time passenger flow prediction method according to claim 1, wherein sample data is preprocessed, specifically comprising:
checking the integrity of input sample data, and checking whether missing values, abnormal values, extreme values and discrete values exist in the input data; if so, performing interpolation by using the mean value;
carrying out characteristic one-hot coding processing on the external factor data;
and carrying out data standardization, regularization and standardization, and scaling the data attribute to a specified range so that the sample data has zero mean and unit variance.
4. The method for short-time passenger flow prediction based on multi-source data fusion input according to claim 1, wherein step 105 is followed by further comprising: and evaluating the passenger flow prediction model, wherein the average absolute error (MAE), the Root Mean Square Error (RMSE) and the average absolute percentage error (MAPE) are used as the performance evaluation indexes of the model.
5. The multi-source data fusion input-based short-time passenger flow prediction method according to claim 1, wherein the spatiotemporal feature extraction in step 103 comprises:
abstracting historical passenger flow data into vectors
The timing features are extracted using a long short term memory network LSTM.
6. The multi-source data fusion input-based short-time passenger flow prediction method of claim 5, wherein the extraction of the timing characteristics using the long-short term memory network LSTM specifically comprises:
a forgetting stage, which processes historical output and current input data through a sigmoid activation function to determine the cell state C at the last momentt-1Keeping the current timeThe information of (a);
in the memory stage, newly added input information of the network at the current moment is determined through the sigmoid activation function, and a new candidate vector is created through the tanh activation function based on historical output and current input data
And an output stage, namely outputting output information determined by a sigmoid activation function, and calculating an output value of the LSTM at the current moment based on the updated cell state.
7. A short-time passenger flow prediction system based on multi-source data fusion input is characterized by comprising:
the data acquisition module is used for acquiring sample data, including space-time characteristic data, external factor characteristic data and recent video data;
the preprocessing module is connected with the data acquisition module and is used for preprocessing the sample data;
the characteristic extraction module is connected with the preprocessing module and used for extracting multi-source data characteristics from the sample data, including the extraction of time-space characteristics, external factor characteristics and video characteristics;
the dimension adjusting module is connected with the feature extracting module and used for adjusting the obtained dimensions of the external factor features and the video features through two full connecting layers to ensure that the dimensions of all the features are uniform;
the multi-source data feature fusion module is connected with the dimension adjusting module and used for carrying out multi-source data feature fusion on the sample data and establishing a passenger flow prediction model for obtaining passenger flow prediction;
and the passenger flow prediction module is connected with the multi-source data characteristic fusion module and used for predicting passenger flow data by the passenger flow prediction module and automatically adjusting the working states of various facilities according to the prediction condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111322660.8A CN114037159A (en) | 2021-11-09 | 2021-11-09 | Short-time passenger flow prediction method and system based on multi-source data fusion input |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111322660.8A CN114037159A (en) | 2021-11-09 | 2021-11-09 | Short-time passenger flow prediction method and system based on multi-source data fusion input |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114037159A true CN114037159A (en) | 2022-02-11 |
Family
ID=80137076
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111322660.8A Pending CN114037159A (en) | 2021-11-09 | 2021-11-09 | Short-time passenger flow prediction method and system based on multi-source data fusion input |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114037159A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115273466A (en) * | 2022-07-14 | 2022-11-01 | 中远海运科技股份有限公司 | Monitoring method and system based on flexible lane management and control algorithm |
-
2021
- 2021-11-09 CN CN202111322660.8A patent/CN114037159A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115273466A (en) * | 2022-07-14 | 2022-11-01 | 中远海运科技股份有限公司 | Monitoring method and system based on flexible lane management and control algorithm |
| CN115273466B (en) * | 2022-07-14 | 2024-01-16 | 中远海运科技股份有限公司 | Monitoring method and system based on flexible lane management and control algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ali et al. | Leveraging spatio-temporal patterns for predicting citywide traffic crowd flows using deep hybrid neural networks | |
| CN111223301B (en) | Traffic flow prediction method based on graph attention convolution network | |
| CN113487066B (en) | Long-time-sequence freight volume prediction method based on multi-attribute enhanced graph convolution-Informer model | |
| CN109117883B (en) | SAR image sea ice classification method and system based on long-time memory network | |
| CN112116080A (en) | CNN-GRU water quality prediction method integrated with attention mechanism | |
| CN109583565A (en) | Forecasting Flood method based on the long memory network in short-term of attention model | |
| Li et al. | Flight delay prediction from spatial and temporal perspective | |
| CN112508265A (en) | Time and activity multi-task prediction method and system for business process management | |
| CN111598325A (en) | Traffic speed prediction method based on hierarchical clustering and hierarchical attention mechanism | |
| Chiang et al. | Hybrid time-series framework for daily-based PM 2.5 forecasting | |
| CN113253363A (en) | Lightning activity path prediction method and system | |
| CN113516304A (en) | Space-time joint prediction method and device for regional pollutants based on space-time graph network | |
| CN114662791A (en) | Long time sequence pm2.5 prediction method and system based on space-time attention | |
| CN113298318A (en) | Novel overload prediction method for distribution transformer | |
| CN114037159A (en) | Short-time passenger flow prediction method and system based on multi-source data fusion input | |
| CN113345236A (en) | Time-space traffic state prediction method based on Transformer network | |
| CN117233869B (en) | Site short-term wind speed prediction method based on GRU-BiTCN | |
| Vijayalakshmi et al. | Multivariate Congestion Prediction using Stacked LSTM Autoencoder based Bidirectional LSTM Model. | |
| CN117494034A (en) | Air quality prediction method based on traffic congestion index and multi-source data fusion | |
| CN117436653A (en) | Prediction model construction method and prediction method for travel demands of network about vehicles | |
| CN117435997A (en) | Elevator early warning and on-demand maintenance method based on transducer and time sequence compression mechanism | |
| CN111524349B (en) | Context feature injected multi-scale traffic flow prediction model establishing method and using method | |
| CN117172461A (en) | Automatic driving bus dispatching system and bus dispatching method based on passenger flow prediction | |
| Wang et al. | Metroeye: A weather-aware system for real-time metro passenger flow prediction | |
| Porto et al. | Machine learning approaches to extreme weather events forecast in urban areas: Challenges and initial results |
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 |