CN116151478A - Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm - Google Patents

Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm Download PDF

Info

Publication number
CN116151478A
CN116151478A CN202310339642.3A CN202310339642A CN116151478A CN 116151478 A CN116151478 A CN 116151478A CN 202310339642 A CN202310339642 A CN 202310339642A CN 116151478 A CN116151478 A CN 116151478A
Authority
CN
China
Prior art keywords
traffic flow
short
flow data
search algorithm
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
Application number
CN202310339642.3A
Other languages
Chinese (zh)
Inventor
卢金勤
朱立
邓芳明
傅清丁
岳康
朱淘淘
罗梓铭
王志强
于景浩
郭锡昭
朱磊
姚道金
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangxi Kingroad Technology Development Co ltd
Original Assignee
Jiangxi Kingroad Technology Development Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangxi Kingroad Technology Development Co ltd filed Critical Jiangxi Kingroad Technology Development Co ltd
Priority to CN202310339642.3A priority Critical patent/CN116151478A/en
Publication of CN116151478A publication Critical patent/CN116151478A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0125—Traffic data processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • 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
    • 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
    • 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
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Educational Administration (AREA)
  • Biophysics (AREA)
  • Operations Research (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Chemical & Material Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Analytical Chemistry (AREA)
  • Primary Health Care (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Quality & Reliability (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention discloses a short-time traffic flow prediction method, equipment and medium for improving a sparrow search algorithm, wherein the method is applied to a short-time traffic flow prediction system comprising a short-time traffic flow prediction model and used for improving the sparrow search algorithm, and comprises the following steps: acquiring traffic flow data of a road to be predicted; and processing the traffic flow data through a short-time traffic flow prediction model to obtain a traffic flow prediction result of the road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time memory neural network training. By the method, when the traffic flow is predicted for a short time, the prediction speed can be increased to a certain extent, and the prediction precision is improved.

Description

Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm
Technical Field
The invention relates to the technical field of computers, in particular to a short-time traffic flow prediction method, short-time traffic flow prediction equipment and short-time traffic flow prediction medium for improving sparrow search algorithm.
Background
With the rapid development of national economy and the general improvement of the living standard of people, the possession of motor vehicles and the traffic of roads are rapidly increased, and a plurality of cities are introduced into urban highway traffic flow prediction systems.
In daily life, prediction of urban highway traffic flow is often affected by factors such as road conditions, time points, weather changes, etc., so that urban highway traffic flow data has high uncertainty and is also of unknown and explicit regularity.
In the prior art, when predicting traffic flow of a road, a traditional long-time memory network method with long-time memory is adopted to train network parameters of the long-time memory network.
At present, a plurality of short-time traffic flow prediction methods, in particular to a neural network related algorithm, are widely applied to various fields. In particular, the development of deep learning greatly enhances the accuracy of neural network calculations. Where short-term traffic flow refers to the number of vehicles passing through a road level in a short period of time (typically taking 5 minutes to 20 minutes).
In the process of realizing the technical scheme of the embodiment of the application, the inventor at least finds that the following technical problems exist in the prior art:
the super parameters in the long-short-term memory network model are difficult to determine, and a traffic flow prediction model is difficult to accurately build by a person skilled in the art, so that the prediction accuracy of traffic flow is reduced.
For the short-time traffic flow prediction method, the response time of the deep learning algorithm is longer from the aspect of the current algorithm, and the traffic flow prediction speed is slow. In practical applications, deep learning algorithms are not well suited for short-term traffic flow predictions.
In conclusion, the existing short-time traffic flow prediction method has the technical problems of low prediction speed and low prediction accuracy.
Disclosure of Invention
The embodiment of the application provides a short-time traffic flow prediction method, device and medium for improving a sparrow search algorithm, which solve the technical problems of low prediction speed and low prediction precision in the existing short-time traffic flow prediction method in the prior art.
In one aspect, an embodiment of the present invention provides a short-term traffic flow prediction method for improving a sparrow search algorithm, which is applied to a short-term traffic flow prediction system for improving the sparrow search algorithm, where the short-term traffic flow prediction system for improving the sparrow search algorithm includes a short-term traffic flow prediction model, and the method includes: acquiring traffic flow data of a road to be predicted; and processing the traffic flow data through the short-time traffic flow prediction model to obtain a traffic flow prediction result of a road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-and-short-term memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules.
Optionally, the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time and short-time memory neural network training, and specifically comprises the following steps: acquiring historical traffic flow data; based on the historical traffic flow data and an improved sparrow search algorithm, obtaining an initialized long-short-term memory neural network super-parameter; and training the super parameters of the initialized long-short-time memory neural network through the long-short-time memory neural network and the historical traffic flow data to obtain the short-time traffic flow prediction model.
Optionally, the obtaining an initialization long-short-term memory neural network hyper-parameter based on the historical traffic flow data and the improved sparrow search algorithm specifically includes: using the historical traffic flow data as population individuals, setting population quantity, initializing the population by using tent chaotic mapping, and calculating the fitness value of each individual; selecting a specific proportion of discoverers from the population individuals according to the fitness value, and updating the positions of the discoverers according to a first preset rule; selecting a specific amount of joiners from the rest population individuals, and updating the positions of the joiners according to a second preset rule; randomly selecting a specific amount of alertors from the population of individuals, and updating the positions of the alertors according to a third preset rule; obtaining an optimal position solution according to a Gaussian variation disturbance strategy; judging whether a termination condition is reached or not according to a greedy rule, and generating a judgment result; and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network.
Optionally, after the obtaining the traffic flow data of the road to be predicted, before the processing the traffic flow data by the short-time traffic flow prediction model, the method further includes the steps of: and correcting the traffic flow data.
Optionally, the correcting the traffic flow data specifically includes: and carrying out data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, restoration and/or normalization processing on the traffic flow data.
Optionally, after the obtaining of the historical traffic flow data and before the obtaining of the initialization long-short-term memory neural network super-parameters, the method further comprises the steps of: correcting the historical traffic flow data.
Optionally, the correcting the historical traffic flow data specifically includes: and carrying out data noise reduction, synchronous extrusion wavelet noise reduction, abnormal data identification and repair and/or normalization on the historical traffic flow data.
In another aspect, embodiments of the present application further provide a computer device including a memory and a processor, the memory storing a computer program, the processor implementing the steps of a short-term traffic flow prediction method that improves a sparrow search algorithm when the computer program is executed.
Embodiments of the present application also provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of a short-term traffic flow prediction method that improves a sparrow search algorithm.
Embodiments of the present application also provide a computer program product comprising a computer program which, when executed by a processor, implements the steps of a short-term traffic flow prediction method that improves a sparrow search algorithm.
One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:
acquiring traffic flow data of a road to be predicted; and processing the traffic flow data through the short-time traffic flow prediction model to obtain a traffic flow prediction result of a road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-and-short-term memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules. The short-time traffic flow prediction model of the long-time memory neural network is generated based on the training of the improved sparrow search algorithm when the short-time traffic flow is predicted, and the improved sparrow search algorithm has the characteristics of higher anti-interference capability and global search capability, higher convergence speed and smoother convergence curve, so that the short-time traffic flow prediction model can be converged to a global optimal solution to a great extent, and the prediction speed can be accelerated to a certain extent and the prediction precision can be improved when the short-time traffic flow is predicted by adopting the short-time traffic flow prediction model generated based on the training of the improved sparrow search algorithm.
Further, the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time and short-time memory neural network training, and specifically comprises the following steps: acquiring historical traffic flow data; based on the historical traffic flow data and an improved sparrow search algorithm, obtaining an initialized long-short-term memory neural network super-parameter; and training the super parameters of the initialized long-short-time memory neural network through the long-short-time memory neural network and the historical traffic flow data to obtain the short-time traffic flow prediction model. When the short-time traffic flow is predicted, the short-time traffic flow prediction model is provided with the short-time traffic flow prediction model, the short-time traffic flow prediction model is provided with the short-time memory neural network super-parameters, and the short-time traffic flow prediction model is provided with the short-time memory neural network super-parameters.
Still further, the obtaining an initialized long-short-term memory neural network hyper-parameter based on the historical traffic flow data and the improved sparrow search algorithm specifically comprises: using the historical traffic flow data as population individuals, setting population quantity, initializing the population by using tent chaotic mapping, and calculating the fitness value of each individual; selecting a specific proportion of discoverers from the population individuals according to the fitness value, and updating the positions of the discoverers according to a first preset rule; selecting a specific amount of joiners from the rest population individuals, and updating the positions of the joiners according to a second preset rule; randomly selecting a specific amount of alertors from the population of individuals, and updating the positions of the alertors according to a third preset rule; obtaining an optimal position solution according to a Gaussian variation disturbance strategy; judging whether a termination condition is reached or not according to a greedy rule, and generating a judgment result; and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network. The initialized long-short-term memory neural network hyper-parameters obtained by adopting the improved sparrow search algorithm are very sensitive to extra-large or extra-small error reaction, and can reflect that the short-term traffic flow prediction model has good stability, so that the accuracy of a traffic flow prediction result is ensured.
Still further, after the obtaining of the traffic flow data of the road to be predicted, before the processing of the traffic flow data by the short-time traffic flow prediction model, the method further includes the steps of: and correcting the traffic flow data. By correcting the traffic flow data, the accuracy of the traffic flow data prediction result can be improved. The accuracy of the traffic flow data, which is taken as the input data of the short-time traffic flow prediction model, affects the output result of the short-time traffic flow prediction model, namely, the more accurate the traffic flow data is, the closer the traffic flow prediction result is to the real traffic flow.
Still further, the correcting the traffic flow data specifically includes: and carrying out data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, repair and/or normalization processing on the traffic flow data, so that a prediction result can be more accurate.
Still further, after the obtaining of the historical traffic flow data and before the obtaining of the initialization long-short-term memory neural network super-parameters, the method further comprises the steps of: correcting the historical traffic flow data. Because the historical traffic flow data is not only used for generating the super parameters of the initialized long-short memory neural network, but also used for obtaining the short-time traffic flow prediction model, the accuracy of the historical traffic flow data has a great influence on the accuracy of the traffic flow prediction result, and the accuracy of the traffic flow data prediction result can be further improved by correcting the historical traffic flow data.
Still further, the correcting the historical traffic flow data specifically includes: and carrying out data noise reduction, synchronous extrusion wavelet noise reduction, abnormal data identification, repair and/or normalization on the historical traffic flow data, so that a prediction result can be more accurate.
Drawings
FIG. 1 is a flow chart of a short-term traffic flow prediction method for improving sparrow search algorithm in an embodiment of the present application;
FIG. 2 is a flowchart of a method for generating a short-term traffic flow prediction model by training an improved sparrow search algorithm according to an embodiment of the present application;
FIG. 3 is a flowchart of a method for obtaining an initialization long-short memory neural network hyper-parameter in an embodiment of the application.
Detailed Description
The embodiment of the application provides a short-time traffic flow prediction method, device and medium for improving a sparrow search algorithm, which solve the technical problems of low prediction speed and low prediction precision in the existing short-time traffic flow prediction method.
The technical scheme of an embodiment of the invention aims to solve the problems, and the general idea is as follows:
acquiring traffic flow data of a road to be predicted; and processing the traffic flow data through a short-time traffic flow prediction model to obtain a traffic flow prediction result of the road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules. The short-time traffic flow prediction model of the long-time memory neural network is generated based on the training of the improved sparrow search algorithm when the short-time traffic flow is predicted, and the improved sparrow search algorithm has the characteristics of higher anti-interference capability and global search capability, higher convergence speed and smoother convergence curve, so that the short-time traffic flow prediction model can be converged to a global optimal solution to a great extent, and the prediction speed can be accelerated to a certain extent and the prediction precision can be improved when the short-time traffic flow is predicted by adopting the short-time traffic flow prediction model generated based on the training of the improved sparrow search algorithm.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments. It will be apparent that the described embodiments of the invention are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment provides a short-time traffic flow prediction method for improving a sparrow search algorithm, which is applied to a short-time traffic flow prediction system for improving the sparrow search algorithm.
Referring to fig. 1, a short-term traffic flow prediction method for improving a sparrow search algorithm according to an embodiment of the present invention will be described in detail.
Step 101: acquiring traffic flow data of a road to be predicted;
step 102: and processing the traffic flow data through a short-time traffic flow prediction model to obtain a traffic flow prediction result of the road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules.
The following will take traffic conditions of the cloud flyway in the nan Chang city as an example, and describe in detail the procedure of the short-time traffic flow prediction method for improving the sparrow search algorithm in the embodiment of the present application.
When the short-time traffic flow prediction system of the improved sparrow search algorithm is started, step 101 is started to be executed: and obtaining traffic flow data of the road to be predicted.
The number of the traffic flow data of the urban intersections to be predicted can be preset to be N, and the historical traffic flow data of each relevant road intersection and the historical traffic flow data of the road intersection to be predicted are used as data set samples, namely, each relevant road intersection with higher degree of correlation with the road intersection to be predicted is found, the historical traffic flow data of the relevant road intersection are used as samples together, and the optimal training sample for training the long-short-term memory neural network is further found, so that the prediction precision of the short-term traffic flow prediction model of the long-short-term memory neural network is higher.
Step 101 is implemented in the following steps: and acquiring traffic flow data of the cloud flyway of Nanchang city and the related intersections of the cloud flyway, wherein the traffic flow data comprises the number of vehicles passing through the cloud flyway, the vehicle speed, the occupancy rate and the like. In the practical application process, the user can customize the time interval for acquiring the traffic flow data, for example, 2 minutes is taken as the time interval for acquiring the traffic flow data.
After the traffic flow data is acquired, execution of step 102 is started: and processing traffic flow data through a short-time traffic flow prediction model to obtain a traffic flow prediction result of a road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules.
Step 102 is implemented in the following manner: the short-time traffic flow prediction system of the improved sparrow search algorithm pre-establishes a short-time traffic flow prediction model of the long-time memory neural network, or generates the short-time traffic flow prediction model of the long-time memory neural network based on the improved sparrow search algorithm and long-time memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules.
And processing the traffic flow data of the Nanchang city cloud flyway through a short-time traffic flow prediction model, and outputting a traffic flow prediction result of the Nanchang city cloud flyway through the short-time traffic flow prediction model.
The short-time traffic flow prediction method for improving sparrow search algorithm relates to data processing, data set division, parameter setting, model selection and construction and the like, and the process is complex. The network error can be reduced by adding the hidden layers, but the hidden layers are possibly overfitted, and meanwhile, the complex model has better expressive force, but the gradient change can be caused by the excessive layers, so that the network training is influenced, and the short-time traffic flow prediction model used by the short-time traffic flow prediction method for improving the sparrow search algorithm is four layers in total.
As shown in fig. 2, in order to improve accuracy of the short-time traffic flow prediction model, further increase prediction speed and further improve prediction accuracy, the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-and-short time memory neural network training, and specifically includes the following steps.
Step S1021: acquiring historical traffic flow data;
step S1022: based on historical traffic flow data and an improved sparrow search algorithm, obtaining an initialized long-short-term memory neural network super-parameter;
step S1023: and training the super parameters of the initialized long-short-time memory neural network through the long-short-time memory neural network and the historical traffic flow data to obtain a short-time traffic flow prediction model.
The method for generating the short-time traffic flow prediction model by training the improved sparrow search algorithm is continuously described in detail by taking the traffic condition of the cloud flyway in Nanchang city as an example.
Before training the short-time traffic flow prediction model of the long-time memory neural network, the input traffic flow data, the length of the historical traffic flow data and the control parameters of the short-time traffic flow prediction model need to be preset.
When the user needs to build or update the short-time traffic flow prediction model, or when the system builds or automatically updates the short-time traffic flow prediction model, the step S1021 is started to be executed: a historical traffic flow data is obtained.
In the specific implementation process, step S1021 is, for example: and acquiring historical traffic flow data from a database of a traffic data control center corresponding to the cloud flyway in Nanchang city. In the practical application process, a user can predict 2786 traffic flow data in total in 1 month and 31 days of the section 2023 as a historical traffic flow for training to generate a short-time traffic flow prediction model, and can also select a part of the 2786 traffic flow data as the historical data. Of course, the user may customize the predicted section, which is not limiting in this application.
After acquiring the history traffic flow data, execution of step S1022 is started: based on the historical traffic flow data and the improved sparrow search algorithm, an initialized long-short-term memory neural network hyper-parameter is obtained.
In the specific implementation process, step S1022 is, for example: the historical traffic flow data is divided into two parts, wherein one part of the historical traffic flow data is used as experimental data, and the other part of the historical traffic flow data is used as test data.
And training experimental data of the historical traffic flow data by adopting an improved sparrow search algorithm to obtain the ultra-parameters of the initialized long-short-term memory neural network.
The improved sparrow search algorithm is a heuristic algorithm, can conveniently and accurately determine the super-parameters of the initialization long-short-term memory neural network, can be a fixed value, and does not need to update the super-parameters of the initialization long-short-term memory neural network after obtaining traffic flow data of a road to be predicted each time. Of course, the super parameters of the initializing long-time memory neural network can be updated according to actual requirements, for example, the super parameters of the initializing long-time memory neural network can be updated within a preset time period, for example, within 8 hours, and the method is not limited.
After generating the initialization long-short-time memory neural network super-parameters, start to execute step S1023: and training the super parameters of the initialized long-short-time memory neural network through the long-short-time memory neural network and the historical traffic flow data to obtain a short-time traffic flow prediction model.
In the specific implementation process, step S1023 includes, for example: and training and initializing the super parameters of the long-short-time memory neural network to construct a short-time traffic flow prediction model of the long-short-time memory neural network for the experimental data of the historical traffic flow data, and predicting whether the test data of the historical traffic flow data accords with the prediction expectation or not through the short-time traffic flow prediction model.
Of course, the above implementation process is merely illustrative, in practical application, the experimental data and the test data may be the same set of historical traffic flow data, or different historical traffic flow data may be used, and the historical traffic flow data may be selected as the experimental data and the test data according to practical situations, which is not limited in this application.
The method for constructing the short-time traffic flow prediction model of the long-time memory neural network comprises the following specific steps of:
and a data processing stage: and carrying out normalization processing on traffic flow data of the road to be predicted, and taking the processed data as input.
Setting model parameters: initial values of the parameters that need to be optimized are set.
Setting a fitness function stage: in the process of training and initializing long-time memory neural network super-parameters, a root mean square error function is used for establishing an adaptability function.
Parameter optimization stage: parameters of the long-short-term memory neural network are optimized by improving a sparrow search algorithm.
As shown in fig. 2 and fig. 3, in order to make the initializing long-short-term memory neural network hyper-parameter very sensitive to the extra-large or extra-small error reaction, the stability of the short-term traffic flow prediction model can be reflected, so as to ensure the accuracy of the traffic flow prediction result, and the step S1022 is based on the historical traffic flow data and improves the sparrow search algorithm, so as to obtain an initializing long-short-term memory neural network hyper-parameter, which specifically comprises the following steps.
Step A1021: using historical traffic flow data as population individuals, setting population quantity, carrying out population initialization by using tent chaotic mapping, and calculating the fitness value of each individual;
step A1022: selecting a specific proportion of discoverers from the population individuals according to the fitness value, and updating the positions of the discoverers according to a first preset rule;
step a1023: selecting a specific amount of joiners from the rest population individuals, and updating the positions of the joiners according to a second preset rule;
Step A1024: randomly selecting a specific amount of alertors from the population of individuals, and updating the positions of the alertors according to a third preset rule;
step A1025: obtaining an optimal position solution according to a Gaussian variation disturbance strategy;
step a1026: judging whether a termination condition is reached or not according to a greedy rule, and generating a judgment result;
step a1027: and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network.
Taking traffic conditions of cloud flyways in Nanchang city as an example, a method for obtaining an initialization long-short-term memory neural network super-parameter based on historical traffic flow data and improved sparrow search algorithm is described in detail.
When the user needs to establish or update the super parameter of the initialization long-time memory neural network, or when the system establishes or automatically updates the super parameter of the initialization long-time memory neural network, the step A1021 is started to be executed: and taking the historical traffic flow data as population individuals, setting population quantity, initializing the population by using tent chaotic mapping, and calculating the fitness value of each individual.
In the specific implementation process, step a1021 is, for example: 2786 historical traffic flow data obtained by a database of a traffic data control center corresponding to the cloud flyway in Nanchang city, 1 month and 31 days in 2023 are used as population individuals in the improved sparrow searching algorithm.
Setting population quantity, carrying out population initialization by using tent chaotic mapping, calculating fitness values of individuals of each population, and obtaining optimal and worst fitness values and positions of individuals of the population from the fitness values, wherein an iterative formula of the tent chaotic mapping is shown as a formula (1):
Figure SMS_1
（1）
where k represents the number of mappings,
Figure SMS_2
the function value of this map at the kth time is shown, < >>
Figure SMS_3
As long as->
Figure SMS_4
Within this desirable range, the system is in a chaotic state, but when +.>
Figure SMS_5
At 0.5, the system willA short period state occurs so we do not generally take 0.5.
After step a1021, step a1022 is started: and selecting a specific proportion of discoverers from the population of individuals according to the fitness value, and updating the positions of the discoverers according to a first preset rule.
Step a1022 is performed, for example: selecting a certain proportion of discoverers from population individuals with good adaptability, updating the positions according to a formula (2), and searching with a dynamic step size;
Figure SMS_6
（2）
wherein,,
Figure SMS_7
is a random number, T represents the current iteration number, T represents the maximum iteration number,
Figure SMS_8
representing the early warning value->
Figure SMS_9
Representing a security value +_>
Figure SMS_10
The weight coefficient is shown as formula (3), Q is a random number which is subject to normal distribution, L is a matrix with dimension of 1×m and the elements inside are all 1, and n population individuals are assumed to form n×m population- >
Figure SMS_11
,
Figure SMS_12
Representing the position of the ith population of individuals in the jth dimension as shown in equation (4).
Figure SMS_13
（3）
Where T represents the current iteration number and T represents the maximum iteration number.
Figure SMS_14
（4）
After step a1022, step a1023 is started: and selecting a specific number of joiners from the rest population individuals, and updating the positions of the joiners according to a second preset rule.
Step a1023 is performed in the specific implementation process, for example: selecting a certain amount of joiners from the rest population individuals, and updating the positions according to a formula (5);
Figure SMS_15
（5）
wherein,,
Figure SMS_16
is the optimal position searched by the finder, +.>
Figure SMS_17
For the current global worst position, A is a matrix with each element being randomly 1 or-1 and the dimension being 1×d, and +.>
Figure SMS_18
。
After step a1023, step a1024 is started: randomly selecting a specific number of alertors from the population of individuals, and updating the positions of the alertors according to a third preset rule.
In the specific implementation process, step a1024 includes, for example: randomly selecting a certain amount of alertors from the population of individuals, and updating the positions according to the formula (6);
Figure SMS_19
（6）
wherein,,
Figure SMS_20
for the current global optimal position, +.>
Figure SMS_21
Is a random number and satisfies the parameters of a control step length of a normal distribution rule with variance of 1 and mean of 0, < > >
Figure SMS_22
Is the fitness of the individuals in the current population, +.>
Figure SMS_23
、
Figure SMS_24
The fitness of the individuals of the population currently globally optimal and globally worst, respectively,/for each individual of the population>
Figure SMS_25
Is an infinitely small constant, +.>
Figure SMS_26
Is a random number.
After step a1024, step a1025 is started: and obtaining an optimal position solution according to the Gaussian variation disturbance strategy.
In the implementation process, step a1025 includes: obtaining a new optimal position solution by using a Gaussian variation disturbance strategy, wherein the new optimal position solution is represented by a formula (7):
Figure SMS_27
（7）
after step a1025, step a1026 is started: and judging whether the termination condition is reached or not according to a greedy rule, and generating a judging result.
Step a1026 is implemented, for example: judging whether a termination condition is reached according to a greedy rule in the formula (8), and generating a judging result;
Figure SMS_28
（8）
wherein,,
Figure SMS_29
indicating the fitness at position x.
After step a1026, step a1027 is started: and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network.
Step a1027 is performed in the specific implementation process, for example: and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network, and ending the program. And when the judgment result indicates that the termination condition is not reached, it is necessary to return to step a1021 again.
The common sparrow searching algorithm uses a random generation method to generate an initial population, when the position is updated, jump type is adopted, and effective control on step length is lacking, so that the global searching capability of the sparrow searching algorithm is poor, and when the optimal position is updated, jump type is adopted, so that the diversity is too low.
The improved sparrow search algorithm uses tent chaotic mapping to enable population initial distribution to be more uniform, global search capacity is enhanced through dynamic step length, and Gaussian disturbance and greedy rules are introduced to enhance optimal feasible solution diversity.
Based on historical traffic flow data and improved sparrow search algorithm, in the method step of obtaining the super-parameters of the initialized long-short-term memory neural network, the initialized parameters comprise population number N, maximum iteration number T, finder proportion PD, scout proportion SD and warning threshold value
Figure SMS_30
。
The method comprises the following specific steps of a long-short-term memory neural network algorithm:
step one, input at the current t moment
Figure SMS_31
And the output of the last moment +.>
Figure SMS_32
For input, equation (9) is used together to calculate amnestic door +.>
Figure SMS_33
The forget gate determines which information from the previous cell needs to be discarded, the closer to zero the output value of equation (9) is between zero and one, the more this information should be discarded, the closer to one the information should be retained.
Figure SMS_34
（9）
Wherein,,
Figure SMS_35
output representing forget gate, +.>
Figure SMS_36
Activating a function for sigmoid->
Figure SMS_37
And->
Figure SMS_38
For the weight matrix of the individual gates,
Figure SMS_39
for biasing the door->
Figure SMS_40
For the input at the current time t +.>
Figure SMS_41
Indicating the output of the last time.
And secondly, transmitting the information of the hidden state of the previous layer and the current input information to an output value of the tanh at the same time to generate a new candidate value vector. Finally the two output values are multiplied together,
Figure SMS_42
which information in the output value of tanh is important and should be preserved, and the output variable +.>
Figure SMS_43
Which information of which should be saved into cell state C.
Figure SMS_44
（10）
Figure SMS_45
（11）
Wherein,,
Figure SMS_48
is the output of the update gate,
Figure SMS_51
、
Figure SMS_54
、
Figure SMS_47
、
Figure SMS_49
Weights in linear relation +.>
Figure SMS_53
And->
Figure SMS_55
Is the bias of each gate ∈>
Figure SMS_46
Is composed of->
Figure SMS_50
A vector created by the function +_>
Figure SMS_52
For the input at the current time t +.>
Figure SMS_56
Indicating the output of the last time.
Step three, using formula (12) to update the information of the memory cell, the information of the memory cell can be obtained from
Figure SMS_57
Update to->
Figure SMS_58
Through the functions of the forgetting gate and the input gate, the information of the storage unit can be selectively modified, and the information of the storage unit at the moment is:
Figure SMS_59
（12）
Wherein,,
Figure SMS_60
is the memory cell state,/->
Figure SMS_61
Is the previous memory cell state.
Step four, the output gate is used to determine the value of the next hidden state, equation (13) is used to calculate the output gate
Figure SMS_62
The value, first, the previous hidden state and the current input data are transferred to +.>
Figure SMS_63
In which the newly acquired state information is transmitted again to
Figure SMS_64
Output sum of->
Figure SMS_65
To determine the information that the hidden state should convey, and to calculate the information of the new state to be output using equation (14).
Figure SMS_66
（13）
Figure SMS_67
（14）/>
Wherein,,
Figure SMS_69
is the output of the output gate,/->
Figure SMS_71
And->
Figure SMS_74
Weights in linear relation +.>
Figure SMS_70
Is biased (is->
Figure SMS_72
Is by
Figure SMS_73
Output of memory cell of layer, +.>
Figure SMS_75
Is the output of the current cell,
Figure SMS_68
Is the output of the last cell.
The root mean square error is the mean value of the square root of the error of the predicted value and the true value, when the predicted value has more outliers, the value is larger, and the response to the extra-large or extra-small error is very sensitive, so that the stability of the model can be reflected, the smaller the value is, the more stable the model is, and the calculation formula is shown as (15):
Figure SMS_76
（15）
where i represents the i-th sample,
Figure SMS_77
is the actual value +.>
Figure SMS_78
For the predicted value, n is the total number of samples and RMSE is the root mean square error.
As shown in fig. 1, in order to improve the accuracy of the traffic flow data prediction result, after the traffic flow data of the road to be predicted is obtained in step 101, before the traffic flow data is processed by the short-time traffic flow prediction model in step 102, the method further includes the steps of: and correcting the traffic flow data. The accuracy of the traffic flow data, which is taken as the input data of the short-time traffic flow prediction model, affects the output result of the short-time traffic flow prediction model, namely, the more accurate the traffic flow data is, the closer the traffic flow prediction result is to the real traffic flow.
A step of correcting traffic flow data, in a specific implementation process, for example: and checking traffic flow data, identifying error data or lost data in the traffic flow data, and correcting the error data or complementing the lost data to correct the traffic flow data.
In order to make the prediction result more accurate, the traffic flow data is corrected, specifically including: and carrying out data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, restoration and/or normalization processing on the traffic flow data. Specific combinations of correction modes can be selected according to actual requirements, and the application is not limited.
As shown in fig. 2, in order to further improve the accuracy of the traffic flow data prediction result, after obtaining the historical traffic flow data in step S1021, before obtaining the initialized long-short-term memory neural network superparameter in step S1022, the method further includes the steps of: the historical traffic flow data is corrected. Because the historical traffic flow data is not only used for generating the super parameters of the initialized long-short-term memory neural network, but also used for obtaining the short-term traffic flow prediction model, the accuracy of the historical traffic flow data has a great influence on the accuracy of the traffic flow prediction result, and the accuracy of the traffic flow data prediction result can be further improved by correcting the historical traffic flow data.
A step of correcting historical traffic flow data, in a specific implementation process, for example: and checking the historical traffic flow data, identifying error data or lost data in the historical traffic flow data, and correcting the error data or complementing the lost data so as to correct the historical traffic flow data.
In order to make the prediction result more accurate, correcting the historical traffic flow data specifically includes: and carrying out data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, restoration and/or normalization processing on the historical traffic flow data. Specific combinations of correction modes can be selected according to actual requirements, and the application is not limited.
Another embodiment of the invention provides a computer device comprising a memory storing a computer program and a processor implementing the steps of a short-term traffic flow prediction method for improving a sparrow search algorithm when the computer program is executed by the processor.
Another embodiment of the present invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of a short-term traffic flow prediction method that improves a sparrow search algorithm.
Another embodiment of the invention provides a computer program product comprising a computer program which, when executed by a processor, performs the steps of a short-term traffic flow prediction method that improves a sparrow search algorithm.
One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:
acquiring traffic flow data of a road to be predicted; and processing the traffic flow data through a short-time traffic flow prediction model to obtain a traffic flow prediction result of the road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules. The short-time traffic flow prediction model of the long-time memory neural network is generated based on the training of the improved sparrow search algorithm when the short-time traffic flow is predicted, and the improved sparrow search algorithm has the characteristics of higher anti-interference capability and global search capability, higher convergence speed and smoother convergence curve, so that the short-time traffic flow prediction model can be converged to a global optimal solution to a great extent, and the prediction speed can be accelerated to a certain extent and the prediction precision can be improved when the short-time traffic flow is predicted by adopting the short-time memory neural network trained by the improved sparrow search algorithm.
Further, the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-time and short-time memory neural network training, and specifically comprises the following steps: acquiring historical traffic flow data; based on historical traffic flow data and an improved sparrow search algorithm, obtaining an initialized long-short-term memory neural network super-parameter; and training the super parameters of the initialized long-short-time memory neural network through the long-short-time memory neural network and the historical traffic flow data to obtain a short-time traffic flow prediction model. When the short-time traffic flow is predicted, the short-time traffic flow prediction model is provided with the short-time traffic flow prediction model, the short-time traffic flow prediction model is provided with the short-time memory neural network super-parameters, and the short-time traffic flow prediction model is provided with the short-time memory neural network super-parameters.
Still further, based on historical traffic flow data and improved sparrow search algorithm, obtaining an initialized long-short-term memory neural network hyper-parameter specifically comprises: using historical traffic flow data as population individuals, setting population quantity, carrying out population initialization by using tent chaotic mapping, and calculating the fitness value of each individual; selecting a specific proportion of discoverers from the population individuals according to the fitness value, and updating the positions of the discoverers according to a first preset rule; selecting a specific amount of joiners from the rest population individuals, and updating the positions of the joiners according to a second preset rule; randomly selecting a specific amount of alertors from the population of individuals, and updating the positions of the alertors according to a third preset rule; obtaining an optimal position solution according to a Gaussian variation disturbance strategy; judging whether a termination condition is reached or not according to a greedy rule, and generating a judgment result; and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network. The initialized long-short-term memory neural network hyper-parameters obtained by adopting the improved sparrow search algorithm are very sensitive to extra-large or extra-small error reaction, and can reflect that the short-term traffic flow prediction model has good stability, so that the accuracy of a traffic flow prediction result is ensured.
Further, after the traffic flow data of the road to be predicted is acquired, before the traffic flow data is processed by the short-time traffic flow prediction model, the method further comprises the steps of: and correcting the traffic flow data. By correcting the traffic flow data, the accuracy of the traffic flow data prediction result can be improved. The accuracy of the traffic flow data, which is taken as the input data of the short-time traffic flow prediction model, affects the output result of the short-time traffic flow prediction model, namely, the more accurate the traffic flow data is, the closer the traffic flow prediction result is to the real traffic flow.
Still further, the correction traffic flow data specifically includes: the traffic flow data is subjected to data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, repair and/or normalization processing, so that a prediction result can be more accurate.
Still further, after obtaining a history traffic stream data, before obtaining an initialization long-short-term memory neural network super parameter, the method further comprises the steps of: the historical traffic flow data is corrected. Because the historical traffic flow data is not only used for generating the super parameters of the initialized long-short-term memory neural network, but also used for obtaining the short-term traffic flow prediction model, the accuracy of the historical traffic flow data has a great influence on the accuracy of the traffic flow prediction result, and the accuracy of the traffic flow data prediction result can be further improved by correcting the historical traffic flow data.
Still further, correcting the historical traffic flow data specifically includes: the historical traffic flow data is subjected to data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, restoration and/or normalization processing, so that a prediction result can be more accurate.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, magnetic disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims (10)

1. A short-term traffic flow prediction method for improving a sparrow search algorithm, which is applied to a short-term traffic flow prediction system for improving the sparrow search algorithm, and is characterized in that the short-term traffic flow prediction system for improving the sparrow search algorithm comprises a short-term traffic flow prediction model, and the method comprises the following steps:
acquiring traffic flow data of a road to be predicted;
and processing the traffic flow data through the short-time traffic flow prediction model to obtain a traffic flow prediction result of a road to be predicted, wherein the short-time traffic flow prediction model is generated based on an improved sparrow search algorithm and long-and-short-term memory neural network training, and the improved sparrow search algorithm adopts tent chaotic mapping, dynamic step length, gaussian disturbance and greedy rules.
2. The method of claim 1, wherein the short-term traffic flow prediction model is generated based on an improved sparrow search algorithm and long-term memory neural network training, and specifically comprises:
acquiring historical traffic flow data;
based on the historical traffic flow data and an improved sparrow search algorithm, obtaining an initialized long-short-term memory neural network super-parameter;
and training the super parameters of the initialized long-short-time memory neural network through the long-short-time memory neural network and the historical traffic flow data to obtain the short-time traffic flow prediction model.
3. The method of claim 2, wherein the obtaining an initialized long-term memory neural network hyper-parameter based on the historical traffic flow data and the modified sparrow search algorithm, comprises:
using the historical traffic flow data as population individuals, setting population quantity, initializing the population by using tent chaotic mapping, and calculating the fitness value of each individual;
selecting a specific proportion of discoverers from the population individuals according to the fitness value, and updating the positions of the discoverers according to a first preset rule;
selecting a specific amount of joiners from the rest population individuals, and updating the positions of the joiners according to a second preset rule;
randomly selecting a specific amount of alertors from the population of individuals, and updating the positions of the alertors according to a third preset rule;
obtaining an optimal position solution according to a Gaussian variation disturbance strategy;
judging whether a termination condition is reached or not according to a greedy rule, and generating a judgment result;
and when the judging result shows that the termination condition is reached, taking the optimal position solution as the super parameter of the initialized long-short-term memory neural network.
4. The method of claim 1, further comprising the step of, after said obtaining traffic flow data for a link to be predicted, prior to said processing said traffic flow data by said short-term traffic flow prediction model:
And correcting the traffic flow data.
5. The method of claim 4, wherein said correcting said traffic flow data comprises:
and carrying out data noise reduction, synchronous extrusion wavelet noise reduction processing, abnormal data identification, restoration and/or normalization processing on the traffic flow data.
6. The method of claim 2, further comprising the step of, after said obtaining a history of traffic flow data, prior to said obtaining an initialization long-duration memory neural network hyper-parameter:
correcting the historical traffic flow data.
7. The method of claim 6, wherein said correcting said historical traffic flow data specifically comprises:
and carrying out data noise reduction, synchronous extrusion wavelet noise reduction, abnormal data identification and repair and/or normalization on the historical traffic flow data.
8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1-7 when the computer program is executed.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1-7.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1-7.
CN202310339642.3A 2023-04-03 2023-04-03 Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm Pending CN116151478A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310339642.3A CN116151478A (en) 2023-04-03 2023-04-03 Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310339642.3A CN116151478A (en) 2023-04-03 2023-04-03 Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm

Publications (1)

Publication Number Publication Date
CN116151478A true CN116151478A (en) 2023-05-23

Family

ID=86350825

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310339642.3A Pending CN116151478A (en) 2023-04-03 2023-04-03 Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm

Country Status (1)

Country Link
CN (1) CN116151478A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118285753A (en) * 2024-04-20 2024-07-05 广东诺凯科技有限公司 Internet intelligent remote blood pressure abnormal data monitoring system based on big data

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020010717A1 (en) * 2018-07-13 2020-01-16 南京理工大学 Short-term traffic flow prediction method based on spatio-temporal correlation
CN112329934A (en) * 2020-11-17 2021-02-05 江苏科技大学 RBF neural network optimization algorithm based on improved sparrow search algorithm
CN114372408A (en) * 2021-12-11 2022-04-19 上海电机学院 Short-term power load prediction method and device based on chaotic sparrow search algorithm
CN115374689A (en) * 2022-01-17 2022-11-22 浙江科技学院 Air quality index prediction method based on improved sparrow search algorithm optimization

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020010717A1 (en) * 2018-07-13 2020-01-16 南京理工大学 Short-term traffic flow prediction method based on spatio-temporal correlation
CN112329934A (en) * 2020-11-17 2021-02-05 江苏科技大学 RBF neural network optimization algorithm based on improved sparrow search algorithm
CN114372408A (en) * 2021-12-11 2022-04-19 上海电机学院 Short-term power load prediction method and device based on chaotic sparrow search algorithm
CN115374689A (en) * 2022-01-17 2022-11-22 浙江科技学院 Air quality index prediction method based on improved sparrow search algorithm optimization

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
孟闯: "道路交通流数据预测方法研究综述", 计算机工程与应用, pages 1 - 6 *
陈玺: "基于ISSA-LSTM的超短期风电功率预测", 中国优秀硕士学位论文全文数据库, pages 17 - 43 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118285753A (en) * 2024-04-20 2024-07-05 广东诺凯科技有限公司 Internet intelligent remote blood pressure abnormal data monitoring system based on big data
CN118285753B (en) * 2024-04-20 2024-10-18 广东诺凯科技有限公司 Internet intelligent remote blood pressure abnormal data monitoring system based on big data

Similar Documents

Publication Publication Date Title
CN109829577B (en) Rail train running state prediction method based on deep neural network structure model
CN113570859A (en) Traffic flow prediction method based on asynchronous space-time expansion graph convolution network
CN112086144B (en) Molecule generation method, device, electronic equipment and storage medium
CN110543978A (en) Traffic flow data prediction method and device based on wavelet neural network
CN113159403B (en) Intersection pedestrian track prediction method and device
CN110444022A (en) The construction method and device of traffic flow data analysis model
CN110889530A (en) Destination prediction method based on recurrent neural network and server
CN116110234B (en) Traffic flow prediction method and device based on artificial intelligence and application of traffic flow prediction method and device
CN107704426A (en) Water level prediction method based on extension wavelet-neural network model
CN116151478A (en) Short-time traffic flow prediction method, device and medium for improving sparrow search algorithm
CN111507499B (en) Method, device and system for constructing model for prediction and testing method
CN115691140B (en) Analysis and prediction method for space-time distribution of automobile charging demand
CN114882713B (en) Signal control method, system, equipment and storage medium based on multiple scenes
CN114818484B (en) Training method of driving environment model and prediction method of driving environment information
CN112733452B (en) Track prediction method, track prediction device, electronic equipment and readable storage medium
CN116048112A (en) Medium-and-long distance combat aircraft track prediction method based on ensemble learning
Cao et al. Research On Regional Traffic Flow Prediction Based On MGCN-WOALSTM
CN113112092A (en) Short-term probability density load prediction method, device, equipment and storage medium
Chen et al. A Bayesian Neural Network‐Based Method to Calibrate Microscopic Traffic Simulators
CN111931994A (en) Short-term load and photovoltaic power prediction method, system, equipment and medium thereof
CN118469045B (en) Permafrost upper limit prediction method, device, equipment and storage medium
CN118052347B (en) Travel time estimation method and system based on travel track sequence of floating car
CN111523090B (en) Number time-varying multi-target tracking method based on Gaussian mixture probability hypothesis density
Wang et al. Research on Short-Term Traffic Flow Forecast Based on Improved Cuckoo Search Algorithm
Cong et al. Traffic flow forecasting by grey model and least squares support vector machine algorithm.

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20230523