CN115049155A - Intelligent park parking decision optimization method and system - Google Patents

Intelligent park parking decision optimization method and system Download PDF

Info

Publication number
CN115049155A
CN115049155A CN202210908844.0A CN202210908844A CN115049155A CN 115049155 A CN115049155 A CN 115049155A CN 202210908844 A CN202210908844 A CN 202210908844A CN 115049155 A CN115049155 A CN 115049155A
Authority
CN
China
Prior art keywords
parking
park
intelligent
model
parks
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202210908844.0A
Other languages
Chinese (zh)
Other versions
CN115049155B (en
Inventor
吴阔
钟娟
胡思思
陈积勇
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.)
Hunan Qianqian Kechuang Co ltd
Original Assignee
Hunan Qianqian Kechuang 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 Hunan Qianqian Kechuang Co ltd filed Critical Hunan Qianqian Kechuang Co ltd
Priority to CN202210908844.0A priority Critical patent/CN115049155B/en
Publication of CN115049155A publication Critical patent/CN115049155A/en
Application granted granted Critical
Publication of CN115049155B publication Critical patent/CN115049155B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/34—Route searching; Route guidance
    • G01C21/3407—Route searching; Route guidance specially adapted for specific applications
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/34—Route searching; Route guidance
    • G01C21/3446—Details of route searching algorithms, e.g. Dijkstra, A*, arc-flags, using precalculated routes
    • 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)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • Automation & Control Theory (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention relates to the technical field of parking decision optimization, and discloses an intelligent park parking decision optimization method and system, wherein the method comprises the following steps: collecting parking influence factors of historical parking data in an intelligent park; constructing a multi-influence factor intelligent park parking quantity regression model; predicting the parking quantity of the parking lot on the current date based on a multi-influence factor smart lot parking quantity regression model, and constructing a parking lot parking feature vector; constructing a neural network parking area decision optimization model based on the parking feature vectors of the parking areas; and training and optimizing the neural network parking park decision optimization model by using a heuristic algorithm, and outputting the optimal parking park of the current vehicle by using the model. The method realizes the prediction of the parking quantity of different parking parks based on the parking influence factors of the historical data, calculates the distance between the vehicle position and the different parking parks by utilizing a heuristic algorithm, improves the optimization speed of a decision optimization model, and quickly obtains an available decision model.

Description

Intelligent park parking decision optimization method and system
Technical Field
The invention relates to the technical field of parking decision optimization, in particular to a parking decision optimization method and system for an intelligent park.
Background
With the continuous expansion of the scale of industrial parks, the problem of difficult parking becomes more and more prominent. Although a large amount of software for helping the travelers to stop the vehicle quickly appears in the market, the parking efficiency of the park users is improved. But does not fully achieve the goal of intelligent parking. Because there are independent backstage systems in different park parking area data, the data in each parking area do not realize intercommunicating yet, will appear the problem that the traveller does not have the parking stall and can't confirm nearest parking area and parking route planning when arriving in the parking area, to this problem, this application provides a wisdom garden parking decision optimization method and system, based on the position of different parking areas in the wisdom garden, the vacant parking car parking space rate that the prediction got and the charging standard, select the parking area of optimum as waiting to park the vehicle.
Disclosure of Invention
In view of the above, the invention provides a smart park parking decision optimization method, which aims to (1) calculate parking influence factors of different parking parks based on historical data, construct a multi-influence factor smart park parking quantity regression model to predict the parking quantity of different parking parks on the current date, so as to obtain the vacant parking space rate of different parking parks on the current date, select a parking park with high vacant parking space rate to park, and assist a decision optimization model to perform parking decision optimization; (2) a neural network parking park decision optimization model is constructed based on parking park parking feature vectors, and the distances between the current position of a vehicle and the positions of different parking parks are calculated by utilizing a heuristic algorithm, so that the model training speed is increased, an available model is quickly obtained, and the optimal parking park position is obtained based on the decision optimization model for parking.
The intelligent park parking decision optimization method provided by the invention comprises the following steps:
s1: collecting parking influence factors of historical parking data in the intelligent park, and preprocessing the collected parking influence factors;
s2: constructing a multi-influence-factor regression model of the parking quantity of the intelligent park, wherein the model takes the preprocessed parking influence factors as input and takes the prediction result of the parking quantity on the current date as output;
s3: predicting the parking quantity of the parking park on the current date based on a multi-influence factor smart park parking quantity regression model, and constructing a parking feature vector of the parking park;
s4: constructing a neural network parking lot decision optimization model based on parking feature vectors of the parking lots, wherein the decision optimization model takes the positions of vehicles and the parking feature vectors as input and takes the optimal parking lot in the intelligent lot as output;
s5: and training and optimizing the neural network parking park decision optimization model by using a heuristic algorithm, inputting the position of the vehicle into the model after training and optimizing, outputting the optimal parking park by using the model, sending the position of the optimal parking park to a vehicle navigation system, and driving the vehicle to reach the optimal parking park by using a driver according to navigation information for parking.
As a further improvement of the method of the invention:
optionally, the step S1 of collecting parking impact factors of historical parking data in the smart park includes:
the intelligent park comprises intelligent parks, intelligent parks and intelligent parks, wherein the intelligent parks include historical parks and parks in the intelligent parks, the historical parks include the historical parks and parks in the historical records data, the historical parks include order number, vehicle license plate number, park position, park start time, park finish time, park expense, date, weather, temperature of park, withdraw the influence factor of parking in different parks from the historical parks;
the parking influence factors comprise parking park positions, weather influence factors, parking quantity influence factors, holiday influence factors and temperature influence factors, wherein the parking influence factors are based on historical data.
Optionally, the step S1 is to pre-process the collected parking impact factors, including:
to arbitrary parking park k in the wisdom garden, carry out the preliminary treatment to the historical weather chronogenesis data of current date, obtain the weather shadow of current dateNoise factor
Figure 216575DEST_PATH_IMAGE001
：
Figure 527471DEST_PATH_IMAGE002
Wherein:
t represents the current date of the day,
Figure 63626DEST_PATH_IMAGE003
indicating the weather on the past i-th day,
Figure 409156DEST_PATH_IMAGE004
，
Figure 497198DEST_PATH_IMAGE005
indicating that the ith day in the past was a rainy day,
Figure 655909DEST_PATH_IMAGE006
indicating that the ith day in the past is a sunny day;
Figure 385968DEST_PATH_IMAGE007
indicating the effect of the weather on the number of parks on the day of park k on the past day i,
Figure 320426DEST_PATH_IMAGE008
；
the intelligent park is rasterized, and the position coordinate of the parking park k in the intelligent park is
Figure 454735DEST_PATH_IMAGE009
；
Marking holiday dates as festivals
Figure 740223DEST_PATH_IMAGE010
Date of non-holidays
Figure 742814DEST_PATH_IMAGE011
The marked parameters are the holiday influence factors after the coding processing;
the temperature at any date is normalized:
Figure 921991DEST_PATH_IMAGE012
wherein:
Figure 617415DEST_PATH_IMAGE013
the smart campus temperature on day i past the current date t,
Figure 124620DEST_PATH_IMAGE014
expressing the temperature of the intelligent park at any date after normalization processing;
Figure 806268DEST_PATH_IMAGE015
indicating a minimum temperature in historical parking data for the intelligent campus,
Figure 715318DEST_PATH_IMAGE016
representing a maximum temperature in historical parking data for the smart park;
taking the intelligent park temperature time sequence data after normalization processing as a temperature influence factor after preprocessing;
and (3) forming time sequence data by the daily maximum parking quantity of the parking park k, and taking the formed time sequence data of the single-day maximum parking quantity as a parking quantity influence factor.
Optionally, the constructing a multiple-influence-factor regression model of the parking number of the intelligent park in the step S2 includes:
for any parking lot k in the intelligent park, constructing a multi-influence-factor intelligent park parking quantity regression model, wherein the model takes the preprocessed parking influence factors as input and takes the prediction result of the parking quantity on the current date as output;
the regression model of the parking quantity of the multi-influence factor intelligent park is as follows:
Figure 316064DEST_PATH_IMAGE017
Figure 205172DEST_PATH_IMAGE018
Figure 549566DEST_PATH_IMAGE019
wherein:
Figure 313122DEST_PATH_IMAGE020
representing the number of parks of the park k on the current date;
Figure 225715DEST_PATH_IMAGE021
representing a white noise sequence;
Figure 707511DEST_PATH_IMAGE022
represents the maximum number of parking lots k on a single day on the ith past day;
Figure 855596DEST_PATH_IMAGE023
is a regression coefficient, wherein
Figure 863872DEST_PATH_IMAGE024
；
c represents the number of parameters in the model, and L represents the time sequence length of the historical parking data of the parking lot k;
substituting the historical parking data of the parking park k into the regression model of the intelligent park parking number with multiple influence factors, solving by using a least square method to obtain a regression coefficient, wherein the solving flow of the regression coefficient is as follows:
s21: respectively constructing the regression coefficient and the parking influence factor of the parking park k into a matrix form, wherein the regression coefficientThe matrix form of the number is
Figure 72000DEST_PATH_IMAGE025
T represents transposition;
the matrix form of the parking impact factors for parking park k is:
Figure 775513DEST_PATH_IMAGE026
Figure 602655DEST_PATH_IMAGE027
wherein:
Figure 340804DEST_PATH_IMAGE028
a parking impact factor representing the L th day in the past of the parking lot k;
Figure 719833DEST_PATH_IMAGE029
representing the historical single-day total parking of the park k,
Figure 802321DEST_PATH_IMAGE030
represents the total amount of single-day parking of the parking park k on the last L-th day;
s22: constructing a loss function of a regression model based on a least square method:
Figure 292208DEST_PATH_IMAGE031
wherein:
Figure 619284DEST_PATH_IMAGE032
traces representing a computational matrix;
s23: calculating partial derivatives of the regression coefficients based on a loss function:
Figure 310159DEST_PATH_IMAGE033
let the left expression be 0, the solution result of the regression coefficient is:
Figure 988265DEST_PATH_IMAGE034
。
optionally, in the step S3, predicting the parking quantity of the parking lot on the current date based on the multiple influence factor smart park parking quantity regression model, and constructing a parking lot parking feature vector, including:
based on the parking quantity of many influence factors wisdom garden parking quantity regression model prediction parking garden in the parking quantity of current date, then to the different parking garages in the wisdom garden, the current date parking quantity collection that the prediction obtained is:
Figure 672056DEST_PATH_IMAGE035
wherein:
Figure 119218DEST_PATH_IMAGE036
the parking number of the parking park K on the current date t is represented, and the K represents the total number of the parking parks in the intelligent park;
and (3) constructing parking characteristic vectors of the parking parks, wherein the parking characteristic vectors of any parking park k are as follows:
Figure 574470DEST_PATH_IMAGE037
wherein:
Figure 739873DEST_PATH_IMAGE038
indicating the location coordinates of the parking park k on the smart park,
Figure 446929DEST_PATH_IMAGE039
；
Figure 279755DEST_PATH_IMAGE040
representing the number of parking spaces in the parking park k;
Figure 905909DEST_PATH_IMAGE041
the result of the normalization process representing the hourly parking charge for the parking park k,
Figure 181776DEST_PATH_IMAGE042
，
Figure 82736DEST_PATH_IMAGE043
represents the maximum hourly parking charge for the intelligent campus,
Figure 238911DEST_PATH_IMAGE044
represents the minimum value of hourly parking charge in the intelligent park,
Figure 176911DEST_PATH_IMAGE045
indicating an hourly parking charge for parking park k.
Optionally, in the step S4, constructing a neural network parking lot decision optimization model based on the parking feature vectors of the parking lot, including:
constructing a neural network parking park decision optimization model based on parking park parking feature vectors, wherein the decision optimization model takes a vehicle position and the parking feature vectors as input and takes an optimal parking park in an intelligent park as output;
the neural network parking area decision optimization model comprises an input layer, a convolution layer and a full connection layer, wherein the input of the input layer is vehicle position coordinates and parking characteristic vectors of different parking areas, and the input represents
Figure 316905DEST_PATH_IMAGE046
Comprises the following steps:
Figure 755977DEST_PATH_IMAGE047
wherein:
Figure 156871DEST_PATH_IMAGE048
is the position coordinates of the vehicle to be parked,
Figure 655985DEST_PATH_IMAGE049
；
Figure 17697DEST_PATH_IMAGE050
the parking characteristic vector of a parking park k in the intelligent park;
the input layer reconstructs the input representation and coordinates the position
Figure 870246DEST_PATH_IMAGE051
Distance of path from parking park position coordinate
Figure 266592DEST_PATH_IMAGE052
Adding the parking feature vector into the parking feature vector of the corresponding parking park, then:
Figure 671029DEST_PATH_IMAGE053
wherein:
Figure 146135DEST_PATH_IMAGE054
the parking characteristic vector of the updated parking park k is obtained;
the output representation of the input layer
Figure 661430DEST_PATH_IMAGE055
Comprises the following steps:
Figure 177862DEST_PATH_IMAGE056
to output the representation
Figure 628566DEST_PATH_IMAGE055
As a convolutional layerIn, convolution layer pair
Figure 230449DEST_PATH_IMAGE055
Convolution is carried out, the characteristic map representation of parking park characteristic vector in the wisdom garden is obtained to in showing the characteristic map and inputing the full connection layer, the full connection layer utilizes the softmax function to output the probability that different parking parks are the best parking park, selects the parking that the probability is the biggest to be the best parking park, and exports the position coordinate of best parking park.
Optionally, in the step S5, training and optimizing the neural network parking lot decision optimization model by using a heuristic algorithm, including:
calculating the path distance between a vehicle to be parked and a parking park in a neural network parking park decision optimization model by using a heuristic algorithm, adding the calculated path distance into a parking feature vector of the parking park to obtain an output representation of an input layer, acquiring a training data set, and performing optimization solution on parameters in the neural network parking park decision optimization model by using a gradient descent algorithm to obtain a neural network parking park decision optimization model after training optimization;
the method comprises the following steps that the format of a training data set is the position of a vehicle to be parked, a parking characteristic vector of a parking park and a determined optimal parking park;
the path distance solving process based on the heuristic algorithm comprises the following steps:
s51: setting the starting position to
Figure 549434DEST_PATH_IMAGE057
The terminal position is any parking park position
Figure 779427DEST_PATH_IMAGE058
Wherein the smart campus has been rasterized into a grid map;
s52: initializing the current iteration number d =0 of the heuristic algorithm, and the maximum iteration number is
Figure 525667DEST_PATH_IMAGE059
After each iterationObtaining the next position coordinate of the vehicle running path, and performing algorithm iteration by taking the next position coordinate as the initial position of the next iteration until the end position is reached;
constructing a fitness function of a heuristic algorithm:
Figure 349266DEST_PATH_IMAGE060
Figure 347309DEST_PATH_IMAGE061
wherein:
Figure 307175DEST_PATH_IMAGE062
for the g-th candidate position coordinate at the d-th iteration,
Figure 489895DEST_PATH_IMAGE063
the abscissa representing the coordinates of the position,
Figure 754037DEST_PATH_IMAGE064
is the ordinate of the position coordinate and,
Figure 680405DEST_PATH_IMAGE065
a fitness function representing the candidate location coordinates,
Figure 494777DEST_PATH_IMAGE066
representing the position coordinate determined after the (d-1) th iteration, namely the initial position in the (d) th iteration;
s53: starting position of the d-th iteration
Figure 723764DEST_PATH_IMAGE067
A plurality of candidate position coordinates are outwards diffused, wherein the initial position of the 0 th iteration is
Figure 256377DEST_PATH_IMAGE068
The number of the diffused candidate position coordinates is a random number between 1 and 4,diffusion radius of L B ；
S54: calculating a fitness function value of the diffused candidate position coordinates, and selecting the candidate position coordinate with the minimum fitness function value as an optimal candidate position coordinate;
connecting the iteration initial position of the current round with the optimal candidate position coordinate, if the connecting line passes through the obstacle grid, selecting the candidate position coordinate with the small fitness function in 8 directions of the current connecting line as the suboptimal candidate position coordinate, connecting the initial position, and repeating the step; if the connecting line does not pass through the obstacle grid, the candidate position coordinates of the connecting line are used as the next position coordinates of the vehicle running path, namely the initial position of the next round of algorithm iteration;
s55: judging whether the parking lot reaches the end position, if so, calculating the current travel path distance, and adding the calculated path distance into the parking feature vector of the parking lot; if the end position is not reached, let d = d +1, return is made to step S53.
Optionally, the step S5 of inputting the vehicle position into the training-optimized model, and the model outputting the optimal parking park and sending the optimal parking park position to the vehicle navigation system includes:
the position coordinates of the vehicle to be parked are input into the neural network parking park decision optimization model after training optimization, the model outputs the optimal parking park and sends the optimal parking park position to the vehicle navigation system, and a driver can drive the vehicle to arrive at the optimal parking park according to navigation information to park.
In order to solve the above problem, the present invention further provides an intelligent park parking decision optimization system, including:
the parking quantity prediction module is used for acquiring parking influence factors of historical parking data in the intelligent park, preprocessing the acquired parking influence factors and constructing a multi-influence factor intelligent park parking quantity regression model;
the characteristic vector extraction device is used for constructing a parking characteristic vector of the parking park;
and the parking decision module is used for constructing a neural network parking park decision optimization model based on the parking feature vectors of the parking parks, training and optimizing the neural network parking park decision optimization model by utilizing a heuristic algorithm, inputting the positions of the vehicles into the trained and optimized model, outputting the optimal parking park by the model and sending the optimal parking park position to the vehicle navigation system.
In order to solve the above problem, the present invention also provides an electronic device, including:
a memory storing at least one instruction; and
and the processor executes the instructions stored in the memory to realize the intelligent park parking decision optimization method.
In order to solve the above problem, the present invention further provides a computer-readable storage medium having at least one instruction stored therein, where the at least one instruction is executed by a processor in an electronic device to implement the intelligent park parking decision optimization method.
Compared with the prior art, the invention provides an intelligent park parking decision optimization method, which has the following advantages:
firstly, the scheme provides a historical data influence factor-based current parking quantity prediction method, and a multi-influence factor smart park parking quantity regression model is constructed for any parking park k in a smart park, wherein the model takes a preprocessed parking influence factor as input and takes a prediction result of the parking quantity on the current date as output; the regression model of the parking quantity of the multi-influence factor intelligent park is as follows:
Figure 986435DEST_PATH_IMAGE069
Figure 779948DEST_PATH_IMAGE070
Figure 304470DEST_PATH_IMAGE071
wherein:
Figure 324379DEST_PATH_IMAGE072
representing the number of parks of the park k on the current date;
Figure 467915DEST_PATH_IMAGE073
representing a white noise sequence;
Figure 256880DEST_PATH_IMAGE074
represents the maximum number of parking lots k on a single day on the ith past day;
Figure 217883DEST_PATH_IMAGE075
is a regression coefficient, wherein
Figure 85607DEST_PATH_IMAGE076
(ii) a c represents the number of parameters in the model, and L represents the time sequence length of the historical parking data of the parking park k; for the regression coefficient of which the model is unknown, substituting the historical parking data of the parking park k into the regression model of the parking number of the intelligent park with multiple influence factors, and solving by using a least square method to obtain the regression coefficient, wherein the solving process of the regression coefficient comprises the following steps: respectively constructing the regression coefficient and the parking influence factor of the parking park k into a matrix form, wherein the matrix form of the regression coefficient is
Figure 891889DEST_PATH_IMAGE077
T represents transposition; the matrix form of the parking impact factors for parking park k is:
Figure 800939DEST_PATH_IMAGE078
Figure 542630DEST_PATH_IMAGE079
wherein:
Figure 537131DEST_PATH_IMAGE080
a parking impact factor representing the L th day in the past of the parking lot k;
Figure 147104DEST_PATH_IMAGE081
representing the historical single-day total parking of the park k,
Figure 35294DEST_PATH_IMAGE082
represents the total amount of single-day parking of the parking park k on the last L-th day; constructing a loss function of a regression model based on a least square method:
Figure 806941DEST_PATH_IMAGE083
wherein:
Figure 554317DEST_PATH_IMAGE084
traces representing a computational matrix; calculating partial derivatives of the regression coefficients based on a loss function:
Figure 577768DEST_PATH_IMAGE033
let the left expression be 0, the solution result of the regression coefficient is:
Figure 195831DEST_PATH_IMAGE085
. Based on the parking quantity of many influence factors wisdom garden parking quantity regression model prediction parking garden in the parking quantity of current date, then to the different parking garages in the wisdom garden, the current date parking quantity collection that the prediction obtained is:
Figure 669538DEST_PATH_IMAGE035
wherein:
Figure 996220DEST_PATH_IMAGE072
the number of parks of parking park K at current date t is shown, and K shows the total number of parking parks in the wisdom garden. Compared withAccording to the traditional scheme, the parking influence factors of different parking lots are calculated based on historical data, a multi-influence factor smart park parking quantity regression model is built to realize prediction of parking quantities of different parking lots on the current date, so that the vacant parking lot parking space rates of different parking lots on the current date are obtained, the parking lots with high vacant parking lot parking space rates are selected to park according to a large tendency, and the parking decision optimization model is assisted to optimize parking decisions.
Meanwhile, the scheme provides a parking park decision optimization model based on a neural network, the neural network parking park decision optimization model comprises an input layer, a convolution layer and a full connection layer, the input of the input layer is vehicle position coordinates and parking characteristic vectors of different parking parks, and the input represents
Figure 682417DEST_PATH_IMAGE046
Comprises the following steps:
Figure 686145DEST_PATH_IMAGE047
wherein:
Figure 674960DEST_PATH_IMAGE086
is the position coordinates of the vehicle to be parked,
Figure 865770DEST_PATH_IMAGE087
；
Figure 621237DEST_PATH_IMAGE088
the parking characteristic vector of a parking park k in the intelligent park; the input layer reconstructs the input representation and coordinates the position
Figure 338526DEST_PATH_IMAGE086
Distance of path from parking park position coordinate
Figure 888456DEST_PATH_IMAGE089
Adding the parking characteristic vector into a corresponding parking park, and then:
Figure 566562DEST_PATH_IMAGE090
wherein:
Figure 735506DEST_PATH_IMAGE091
the updated parking characteristic vector of the parking park k is obtained; the output representation of the input layer
Figure 182668DEST_PATH_IMAGE092
Comprises the following steps:
Figure 903499DEST_PATH_IMAGE056
to output the representation
Figure 960579DEST_PATH_IMAGE092
As input for convolutional layers, convolutional layer pairs
Figure 792269DEST_PATH_IMAGE092
And carrying out convolution processing to obtain a characteristic map representation of parking park characteristic vectors in the intelligent park, inputting the characteristic map representation into a full connection layer, outputting the probability of different parking parks as the optimal parking park by utilizing a softmax function in the full connection layer, selecting the parking with the maximum probability as the optimal parking park, and outputting the position coordinate of the optimal parking park. Compared with the traditional scheme, the method has the advantages that the heuristic algorithm is utilized to quickly calculate the path distance between the vehicle to be parked and the parking park in the neural network parking park decision optimization model, the calculated path distance is added into the parking feature vector of the parking park to obtain the output representation of an input layer, the training data set is obtained, the gradient descent algorithm is utilized to carry out optimization solution on the parameters in the neural network parking park decision optimization model to obtain the neural network parking park decision optimization model after training optimization; the path distance solving process based on the heuristic algorithm comprises the following steps: setting the starting position to
Figure 93938DEST_PATH_IMAGE087
End point positionIs set as any parking park position
Figure 126616DEST_PATH_IMAGE093
Wherein the smart campus has been rasterized into a grid map; initializing current iteration number of heuristic algorithm
Figure 779314DEST_PATH_IMAGE094
The maximum number of iterations is
Figure 414695DEST_PATH_IMAGE095
Obtaining the next position coordinate of the vehicle running path after each iteration, and performing algorithm iteration by taking the next position coordinate as the initial position of the next iteration until the final position is reached; constructing a fitness function of a heuristic algorithm:
Figure 961082DEST_PATH_IMAGE060
Figure 23716DEST_PATH_IMAGE061
wherein:
Figure 163711DEST_PATH_IMAGE062
for the g-th candidate position coordinate at the d-th iteration,
Figure 478148DEST_PATH_IMAGE063
the abscissa representing the coordinates of the position,
Figure 754409DEST_PATH_IMAGE064
is the ordinate of the position coordinate and,
Figure 987944DEST_PATH_IMAGE065
a fitness function representing the coordinates of the candidate location,
Figure 244263DEST_PATH_IMAGE066
indicating the determination after the d-1 th iterationI.e. the starting position in the d-th iteration; starting position of the d-th iteration
Figure 955867DEST_PATH_IMAGE067
A plurality of candidate position coordinates are outwards diffused, wherein the initial position of the 0 th iteration is
Figure 352213DEST_PATH_IMAGE068
The number of the diffused candidate position coordinates is a random number between 1 and 4, and the diffusion radius is L B (ii) a Calculating a fitness function value of the diffused candidate position coordinates, and selecting the candidate position coordinate with the minimum fitness function value as an optimal candidate position coordinate; connecting the iteration initial position of the current round with the optimal candidate position coordinate, if the connecting line passes through the obstacle grid, selecting the candidate position coordinate with the small fitness function in 8 directions of the current connecting line as the suboptimal candidate position coordinate, connecting the initial position, and repeating the step; if the connecting line does not pass through the obstacle grid, the candidate position coordinates of the connecting line are used as the next position coordinates of the vehicle running path, namely the initial position of the next round of algorithm iteration; judging whether the parking lot reaches the end position, if so, calculating the current travel path distance, and adding the calculated path distance into the parking feature vector of the parking lot; if the end position is not reached, let it
Figure 897595DEST_PATH_IMAGE096
And returning to the step. Compared with the traditional scheme, the method and the device have the advantages that the neural network parking park decision optimization model is built based on the parking feature vectors of the parking parks, the distances between the current position of the vehicle and the positions of different parking parks are calculated by utilizing a heuristic algorithm, so that the model training speed is increased, the available model is quickly obtained, and the optimal parking park position is obtained based on the decision optimization model for parking.
Drawings
FIG. 1 is a schematic flow chart illustrating a parking decision optimization method for an intelligent park according to an embodiment of the present invention;
FIG. 2 is a functional block diagram of an intelligent park parking decision optimization system according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an electronic device for implementing an intelligent park parking decision optimization method according to an embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The embodiment of the application provides an intelligent park parking decision optimization method. The execution subject of the intelligent park parking decision optimization method includes, but is not limited to, at least one of the electronic devices of a server, a terminal and the like that can be configured to execute the method provided by the embodiments of the present application. In other words, the intelligent park parking decision optimization method can be executed by software or hardware installed in a terminal device or a server device, and the software can be a block chain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
Example 1:
s1: gather the parking influence factor of historical parking data in the wisdom garden to carry out the preliminary treatment to the parking influence factor who gathers.
Gather the parking influence factor of historical parking data in the wisdom garden in the S1 step, include:
the intelligent park comprises intelligent parks, intelligent parks and intelligent parks, wherein the intelligent parks include historical parks and parks in the intelligent parks, the historical parks include the historical parks and parks in the historical records data, the historical parks include order number, vehicle license plate number, park position, park start time, park finish time, park expense, date, weather, temperature of park, withdraw the influence factor of parking in different parks from the historical parks;
the parking influence factors comprise parking park positions, weather influence factors, parking quantity influence factors, holiday influence factors and temperature influence factors, wherein the parking influence factors are based on historical data.
The step S1 is to pre-process the collected parking impact factors, including:
to any parking lot k in the intelligent park, preprocessing historical weather time sequence data of the current date to obtain weather influence factors of the current date
Figure 746603DEST_PATH_IMAGE001
：
Figure 261898DEST_PATH_IMAGE002
Wherein:
t represents the current date of the day,
Figure 902963DEST_PATH_IMAGE003
indicating the weather on the past i-th day,
Figure 478301DEST_PATH_IMAGE004
，
Figure 814605DEST_PATH_IMAGE005
indicating that the ith day in the past was a rainy day,
Figure 274536DEST_PATH_IMAGE006
indicating that the ith day in the past is a sunny day;
Figure 379895DEST_PATH_IMAGE007
indicating the effect of the weather on the number of parks on the day of park k on the past day i,
Figure 126134DEST_PATH_IMAGE008
；
the intelligent park is rasterized, and the position coordinate of the parking park k in the intelligent park is
Figure 575832DEST_PATH_IMAGE009
；
Marking holiday dates as festivals
Figure 698509DEST_PATH_IMAGE010
Date of non-holidays
Figure 658375DEST_PATH_IMAGE011
The marked parameters are the holiday influence factors after the coding processing;
the temperature of any date is normalized:
Figure 575515DEST_PATH_IMAGE012
wherein:
Figure 27356DEST_PATH_IMAGE013
the smart campus temperature on day i past the current date t,
Figure 953724DEST_PATH_IMAGE014
expressing the temperature of the intelligent park at any date after normalization processing;
Figure 892730DEST_PATH_IMAGE015
indicating a minimum temperature in historical parking data for the intelligent campus,
Figure 246351DEST_PATH_IMAGE016
representing a maximum temperature in historical parking data for the smart park;
taking the intelligent park temperature time sequence data after normalization processing as a temperature influence factor after preprocessing;
the maximum daily parking number of the parking park k forms time sequence data, the formed time sequence data of the maximum daily parking number is used as a parking number influence factor, and it needs to be explained that the single-day parking number represents the simultaneous parking number at any time of the day.
S2: and constructing a multi-influence-factor regression model of the parking quantity of the intelligent park, wherein the model takes the preprocessed parking influence factors as input and takes the prediction result of the parking quantity on the current date as output.
The step S2 is implemented with a regression model of the parking number of the intelligent park with multiple influence factors, comprising the following steps:
for any parking lot k in the intelligent park, constructing a multi-influence-factor intelligent park parking quantity regression model, wherein the model takes the preprocessed parking influence factors as input and takes the prediction result of the parking quantity on the current date as output;
the regression model of the parking quantity of the intelligent park with multiple influence factors is as follows:
Figure 778964DEST_PATH_IMAGE017
Figure 118809DEST_PATH_IMAGE018
Figure 53267DEST_PATH_IMAGE019
wherein:
Figure 843369DEST_PATH_IMAGE020
representing the number of parks of the park k on the current date;
Figure 220867DEST_PATH_IMAGE021
representing a white noise sequence;
Figure 754616DEST_PATH_IMAGE022
represents the maximum number of parking lots k on a single day on the ith past day;
Figure 543581DEST_PATH_IMAGE023
is a regression coefficient, wherein
Figure 848791DEST_PATH_IMAGE024
；
c represents the number of parameters in the model, and L represents the time sequence length of the historical parking data of the parking park k;
substituting the historical parking data of the parking park k into the regression model of the intelligent park parking number with multiple influence factors, solving by using a least square method to obtain a regression coefficient, wherein the solving flow of the regression coefficient is as follows:
s21: respectively constructing the regression coefficient and the parking influence factor of the parking park k into a matrix form, wherein the matrix form of the regression coefficient is
Figure 621575DEST_PATH_IMAGE025
T represents transposition;
the matrix form of the parking impact factors for parking park k is:
Figure 427857DEST_PATH_IMAGE026
Figure 195962DEST_PATH_IMAGE027
wherein:
Figure 327866DEST_PATH_IMAGE028
a parking impact factor representing the L th day in the past of the parking lot k;
Figure 56788DEST_PATH_IMAGE029
representing the historical single-day parking amount for the parking lot k,
Figure 807706DEST_PATH_IMAGE030
represents the total amount of single-day parking of the parking park k on the last L-th day;
s22: constructing a loss function of a regression model based on a least square method:
Figure 571263DEST_PATH_IMAGE031
wherein:
Figure 342910DEST_PATH_IMAGE032
traces representing a computational matrix;
s23: calculating partial derivatives of the regression coefficients based on a loss function:
Figure 450805DEST_PATH_IMAGE033
let the left expression be 0, the solution result of the regression coefficient is:
Figure 864469DEST_PATH_IMAGE034
。
s3: and predicting the parking quantity of the parking park on the current date based on the multi-influence factor smart park parking quantity regression model, and constructing parking feature vectors of the parking park.
The quantity of parkking in the parking garden at current date is predicated based on many influence factors wisdom garden parking quantity regression model in the S3 step to construct the parking garden and park the eigenvector, include:
based on the parking quantity of many influence factors wisdom garden parking quantity regression model prediction parking garden in the parking quantity of current date, then to the different parking garages in the wisdom garden, the current date parking quantity collection that the prediction obtained is:
Figure 482532DEST_PATH_IMAGE035
wherein:
Figure 300447DEST_PATH_IMAGE020
the parking number of the parking lot K on the current date t is represented, and the K represents the total number of the parking lots in the intelligent park;
and (3) constructing parking characteristic vectors of the parking parks, wherein the parking characteristic vectors of any parking park k are as follows:
Figure 269540DEST_PATH_IMAGE037
wherein:
Figure 221315DEST_PATH_IMAGE038
indicating the location coordinates of the parking park k on the smart park,
Figure 84098DEST_PATH_IMAGE039
；
Figure 463127DEST_PATH_IMAGE040
representing the number of parking spaces in the parking park k;
Figure 653936DEST_PATH_IMAGE041
the result of the normalization process representing the hourly parking charge for the parking park k,
Figure 19190DEST_PATH_IMAGE042
，
Figure 877424DEST_PATH_IMAGE043
represents the maximum hourly parking charge for the intelligent campus,
Figure 427354DEST_PATH_IMAGE044
represents the minimum hourly parking charge for the intelligent campus,
Figure 457191DEST_PATH_IMAGE045
indicating an hourly parking charge for parking park k.
S4: and constructing a neural network parking park decision optimization model based on the parking feature vectors of the parking parks, wherein the decision optimization model takes the vehicle position and the parking feature vectors as input and takes the optimal parking park in the intelligent park as output.
In the step S4, a neural network parking lot decision optimization model is constructed based on the parking feature vectors of the parking lot, including:
constructing a neural network parking park decision optimization model based on parking park parking feature vectors, wherein the decision optimization model takes a vehicle position and the parking feature vectors as input and takes an optimal parking park in an intelligent park as output;
the neural network parking area decision optimization model comprises an input layer, a convolutional layer and a full connection layer, wherein the input of the input layer is vehicle position coordinates and parking characteristic vectors of different parking areas, and the input represents
Figure 16348DEST_PATH_IMAGE046
Comprises the following steps:
Figure 463510DEST_PATH_IMAGE047
wherein:
Figure 59708DEST_PATH_IMAGE048
is the position coordinates of the vehicle to be parked,
Figure 490689DEST_PATH_IMAGE049
；
Figure 322379DEST_PATH_IMAGE050
the parking characteristic vector of a parking park k in the intelligent park;
the input layer reconstructs the input representation and coordinates the position
Figure 748681DEST_PATH_IMAGE051
Distance of path from parking park position coordinate
Figure 640413DEST_PATH_IMAGE052
Parking feature vector added to corresponding parking parkIn the step (1), then:
Figure 293112DEST_PATH_IMAGE053
wherein:
Figure 803859DEST_PATH_IMAGE054
the parking characteristic vector of the updated parking park k is obtained;
the output representation of the input layer
Figure 491192DEST_PATH_IMAGE055
Comprises the following steps:
Figure 553826DEST_PATH_IMAGE056
to output the representation
Figure 54339DEST_PATH_IMAGE055
As input for convolutional layers, convolutional layer pairs
Figure 493411DEST_PATH_IMAGE055
And carrying out convolution processing to obtain a characteristic map representation of parking park characteristic vectors in the intelligent park, inputting the characteristic map representation into a full connection layer, outputting the probability of different parking parks as the optimal parking park by utilizing a softmax function in the full connection layer, selecting the parking with the maximum probability as the optimal parking park, and outputting the position coordinate of the optimal parking park.
S5: and training and optimizing the neural network parking park decision optimization model by using a heuristic algorithm, inputting the position of the vehicle into the model after training and optimizing, outputting the optimal parking park by the model, sending the position of the optimal parking park to a vehicle navigation system, and driving the vehicle to reach the optimal parking park by a driver according to navigation information for parking.
In the step S5, a heuristic algorithm is used to train and optimize the neural network parking lot decision optimization model, including:
calculating the path distance between a vehicle to be parked and a parking park in a neural network parking park decision optimization model by using a heuristic algorithm, adding the calculated path distance into a parking feature vector of the parking park to obtain an output representation of an input layer, acquiring a training data set, and performing optimization solution on parameters in the neural network parking park decision optimization model by using a gradient descent algorithm to obtain a neural network parking park decision optimization model after training optimization;
the method comprises the following steps that the format of a training data set is the position of a vehicle to be parked, a parking characteristic vector of a parking park and a determined optimal parking park;
the path distance solving process based on the heuristic algorithm comprises the following steps:
s51: setting the starting position to
Figure 769672DEST_PATH_IMAGE057
The terminal position is any parking park position
Figure 878573DEST_PATH_IMAGE058
Wherein the smart campus has been rasterized into a grid map;
s52: initializing the current iteration number d =0 of the heuristic algorithm, and the maximum iteration number is
Figure 505863DEST_PATH_IMAGE059
Obtaining the next position coordinate of the vehicle running path after each iteration, and performing algorithm iteration by taking the next position coordinate as the initial position of the next iteration until the final position is reached;
constructing a fitness function of a heuristic algorithm:
Figure 483047DEST_PATH_IMAGE060
Figure 4027DEST_PATH_IMAGE061
wherein:
Figure 674043DEST_PATH_IMAGE062
for the g-th candidate position coordinate at the d-th iteration,
Figure 257471DEST_PATH_IMAGE063
the abscissa representing the coordinates of the position,
Figure 913711DEST_PATH_IMAGE064
is the ordinate of the position coordinate and,
Figure 430143DEST_PATH_IMAGE065
a fitness function representing the candidate location coordinates,
Figure 5481DEST_PATH_IMAGE066
representing the position coordinate determined after the (d-1) th iteration, namely the initial position in the (d) th iteration;
s53: starting position of the d-th iteration
Figure 699374DEST_PATH_IMAGE067
A plurality of candidate position coordinates are outwards diffused, wherein the initial position of the 0 th iteration is
Figure 283939DEST_PATH_IMAGE068
The number of the diffused candidate position coordinates is a random number between 1 and 4, and the diffusion radius is L B ；
S54: calculating a fitness function value of the diffused candidate position coordinates, and selecting the candidate position coordinate with the minimum fitness function value as an optimal candidate position coordinate;
connecting the iteration initial position of the current round with the optimal candidate position coordinate, if the connecting line passes through the obstacle grid, selecting the candidate position coordinate with the small fitness function in 8 directions of the current connecting line as the suboptimal candidate position coordinate, connecting the initial position, and repeating the step; if the connecting line does not pass through the obstacle grid, the candidate position coordinates of the connecting line are used as the position coordinates of the next position of the vehicle running path, namely the initial position of the next round of algorithm iteration;
s55: judging whether the parking lot reaches the end position, if so, calculating the current travel path distance, and adding the calculated path distance into the parking feature vector of the parking lot; if the end position is not reached, let d = d +1, return is made to step S53.
In the step S5, inputting the vehicle position into the model after training optimization, and the model outputs the optimal parking park and sends the optimal parking park position to the vehicle navigation system, including:
the position coordinates of the vehicle to be parked are input into the neural network parking park decision optimization model after training optimization, the model outputs the optimal parking park and sends the optimal parking park position to the vehicle navigation system, and a driver can drive the vehicle to arrive at the optimal parking park according to navigation information to park.
Example 2:
fig. 2 is a functional block diagram of an intelligent park parking decision optimization system according to an embodiment of the present invention, which can implement the intelligent park parking decision optimization method in embodiment 1.
The intelligent park parking decision optimization system 100 of the present invention may be installed in an electronic device. According to the realized function, the intelligent park parking decision optimization system can comprise a parking quantity prediction module 101, a feature vector extraction device 102 and a parking decision module 103. The module of the present invention, which may also be referred to as a unit, refers to a series of computer program segments that can be executed by a processor of an electronic device and that can perform a fixed function, and that are stored in a memory of the electronic device.
The parking quantity prediction module 101 is used for acquiring parking influence factors of historical parking data in the intelligent park, preprocessing the acquired parking influence factors and constructing a multi-influence factor intelligent park parking quantity regression model;
the characteristic vector extraction device 102 is used for constructing parking characteristic vectors of the parking park;
the parking decision module 103 is used for constructing a neural network parking park decision optimization model based on the parking feature vectors of the parking parks, training and optimizing the neural network parking park decision optimization model by using a heuristic algorithm, inputting the positions of the vehicles into the trained and optimized model, outputting the optimal parking park by the model and sending the optimal parking park position to a vehicle navigation system.
In detail, in the embodiment of the present invention, when the modules in the intelligent park parking decision optimizing system 100 are used, the same technical means as the intelligent park parking decision optimizing method described in fig. 1 are adopted, and the same technical effect can be produced, which is not described herein again.
Example 3:
fig. 3 is a schematic structural diagram of an electronic device for implementing an intelligent park parking decision optimization method according to an embodiment of the present invention.
The electronic device 1 may comprise a processor 10, a memory 11 and a bus 12, and may further comprise a computer program, such as a smart park parking decision optimization program, stored in the memory 11 and executable on the processor 10.
The memory 11 includes at least one type of readable storage medium, which includes flash memory, removable hard disk, multimedia card, card type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device 1, such as a removable hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the electronic device 1. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device 1. The memory 11 may be used not only to store application software installed in the electronic device 1 and various types of data, such as codes of a parking decision optimization program of a smart park, but also to temporarily store data that has been output or is to be output.
The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same or different functions, including one or more Central Processing Units (CPUs), microprocessors, digital Processing chips, graphics processors, and combinations of various control chips. The processor 10 is a Control Unit of the electronic device, connects various components of the electronic device by using various interfaces and lines, and executes various functions and processes data of the electronic device 1 by operating or executing programs or modules (smart park parking decision optimization programs, etc.) stored in the memory 11 and calling data stored in the memory 11.
The bus 12 may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is arranged to enable connection communication between the memory 11 and at least one processor 10 or the like.
Fig. 3 shows only an electronic device with components, and it will be understood by those skilled in the art that the structure shown in fig. 3 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than those shown, or some components may be combined, or a different arrangement of components.
For example, although not shown, the electronic device 1 may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so as to implement functions of charge management, discharge management, power consumption management, and the like through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device 1 may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.
Further, the electronic device 1 may further include a communication interface 13, and optionally, the communication interface 13 may include a wired interface and/or a wireless interface (such as a WI-FI interface, a bluetooth interface, etc.), which are generally used for establishing a communication connection between the electronic device 1 and other electronic devices.
Optionally, the electronic device 1 may further comprise a user interface, which may be a Display (Display), an input unit (such as a Keyboard), and optionally a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, is suitable for displaying information processed in the electronic device 1 and for displaying a visualized user interface, among other things.
It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.
The intelligent park parking decision optimization program stored in the memory 11 of the electronic device 1 is a combination of instructions that, when executed in the processor 10, enable:
collecting parking influence factors of historical parking data in the intelligent park, and preprocessing the collected parking influence factors;
constructing a multi-influence factor intelligent park parking quantity regression model;
predicting the parking quantity of the parking park on the current date based on a multi-influence factor smart park parking quantity regression model, and constructing a parking feature vector of the parking park;
constructing a neural network parking park decision optimization model based on the parking feature vectors of the parking parks;
and training and optimizing the neural network parking park decision optimization model by using a heuristic algorithm, inputting the position of the vehicle into the model after training and optimizing, outputting the optimal parking park by using the model, sending the position of the optimal parking park to a vehicle navigation system, and driving the vehicle to reach the optimal parking park by using a driver according to navigation information for parking.
Specifically, the specific implementation method of the processor 10 for the instruction may refer to the description of the relevant steps in the embodiments corresponding to fig. 1 to fig. 3, which is not repeated herein.
It should be noted that the above-mentioned numbers of the embodiments of the present invention are merely for description, and do not represent the merits of the embodiments. And the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, apparatus, article, or method that includes the element.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims (8)

1. A smart park parking decision optimization method, the method comprising:
s1: collecting parking influence factors of historical parking data in an intelligent park, and preprocessing the collected parking influence factors;
s2: constructing a multi-influence-factor regression model of the parking quantity of the intelligent park, wherein the model takes the preprocessed parking influence factors as input and takes the prediction result of the parking quantity on the current date as output;
s3: predicting the parking quantity of the parking park on the current date based on a multi-influence factor smart park parking quantity regression model, and constructing a parking feature vector of the parking park;
s4: constructing a neural network parking park decision optimization model based on parking park parking feature vectors, wherein the decision optimization model takes a vehicle position and the parking feature vectors as input and takes an optimal parking park in an intelligent park as output;
s5: the neural network parking park decision optimization model is trained and optimized by using a heuristic algorithm, the position of a vehicle is input into the model after training and optimization, the model outputs an optimal parking park and sends the position of the optimal parking park to a vehicle navigation system, and a driver can drive the vehicle to reach the optimal parking park to park according to navigation information, wherein the neural network parking park decision optimization model is trained and optimized by using the heuristic algorithm, and the method comprises the following steps:
calculating the path distance between a vehicle to be parked and a parking park in a neural network parking park decision optimization model by using a heuristic algorithm, adding the calculated path distance into a parking feature vector of the parking park to obtain an output representation of an input layer, acquiring a training data set, and performing optimization solution on parameters in the neural network parking park decision optimization model by using a gradient descent algorithm to obtain a neural network parking park decision optimization model after training optimization; the path distance solving process based on the heuristic algorithm comprises the following steps:
s51: setting the starting position to
Figure 348679DEST_PATH_IMAGE001
End position ofFor any parking lot position
Figure 360498DEST_PATH_IMAGE002
Wherein the smart campus has been rasterized into a grid map;
s52: initializing the current iteration number d =0 of the heuristic algorithm, and the maximum iteration number is
Figure 810196DEST_PATH_IMAGE003
Obtaining the next position coordinate of the vehicle running path after each iteration, and performing algorithm iteration by taking the next position coordinate as the initial position of the next iteration until the final position is reached;
constructing a fitness function of a heuristic algorithm:
Figure 667294DEST_PATH_IMAGE004
Figure 158318DEST_PATH_IMAGE005
wherein:
Figure 950824DEST_PATH_IMAGE006
for the g-th candidate position coordinate at the d-th iteration,
Figure 996141DEST_PATH_IMAGE007
the abscissa representing the coordinates of the position,
Figure 922508DEST_PATH_IMAGE008
is the ordinate of the position coordinate and,
Figure 127094DEST_PATH_IMAGE009
a fitness function representing the candidate location coordinates,
Figure 215135DEST_PATH_IMAGE010
representing the position coordinate determined after the (d-1) th iteration, namely the initial position in the (d) th iteration;
s53: starting position of the d-th iteration
Figure 13327DEST_PATH_IMAGE011
A plurality of candidate position coordinates are outwards diffused, wherein the initial position of the 0 th iteration is
Figure 353173DEST_PATH_IMAGE012
The number of the diffused candidate position coordinates is a random number between 1 and 4, and the diffusion radius is L B ；
S54: calculating a fitness function value of the diffused candidate position coordinates, and selecting the candidate position coordinate with the minimum fitness function value as an optimal candidate position coordinate;
connecting the iteration initial position of the current round with the optimal candidate position coordinate, if the connecting line passes through the obstacle grid, selecting the candidate position coordinate with the small fitness function in 8 directions of the current connecting line as the suboptimal candidate position coordinate, connecting the initial position, and repeating the step; if the connecting line does not pass through the obstacle grid, the candidate position coordinates of the connecting line are used as the position coordinates of the next position of the vehicle running path, namely the initial position of the next round of algorithm iteration;
s55: judging whether the parking lot reaches the end position, if so, calculating the current travel path distance, and adding the calculated path distance into the parking feature vector of the parking lot; if the end position is not reached, let d = d +1, return is made to step S53.
2. The intelligent park parking decision optimizing method of claim 1 wherein the step of collecting parking impact factors of historical parking data in the intelligent park at S1 comprises:
the intelligent park comprises intelligent parks, intelligent parks and intelligent parks, wherein the intelligent parks include historical parks and parks in the intelligent parks, the historical parks include the historical parks and parks in the historical records data, the historical parks include order number, vehicle license plate number, park position, park start time, park finish time, park expense, date, weather, temperature of park, withdraw the influence factor of parking in different parks from the historical parks;
the parking influence factors comprise parking park positions, weather influence factors, parking quantity influence factors, holiday influence factors and temperature influence factors, wherein the parking influence factors are based on historical data.
3. The intelligent park parking decision optimizing method of claim 2 wherein the step of S1 preprocessing the collected parking impact factors includes:
to any parking park k in the smart park, the historical weather time sequence data of the current date is preprocessed to obtain the weather influence factor of the current date
Figure 287631DEST_PATH_IMAGE013
：
Figure 812153DEST_PATH_IMAGE014
Wherein:
t represents the current date of the day,
Figure 720810DEST_PATH_IMAGE015
indicating the weather on the past i-th day,
Figure 988980DEST_PATH_IMAGE016
，
Figure 777945DEST_PATH_IMAGE017
indicating that the last ith day was a rainy day,
Figure 83155DEST_PATH_IMAGE018
indicating that the ith day in the past is a sunny day;
Figure 855939DEST_PATH_IMAGE019
indicating the effect of the weather on the number of parks on the day of park k on the past day i,
Figure 662221DEST_PATH_IMAGE020
；
the intelligent park is rasterized, and the position coordinate of the parking park k in the intelligent park is
Figure 430326DEST_PATH_IMAGE021
；
Marking holiday dates as festivals
Figure 562230DEST_PATH_IMAGE022
Date of non-holidays
Figure 291151DEST_PATH_IMAGE023
The marked parameters are the holiday influence factors after the coding processing;
the temperature of any date is normalized:
Figure 42070DEST_PATH_IMAGE024
wherein:
Figure 540047DEST_PATH_IMAGE025
the smart campus temperature on day i past the current date t,
Figure 577273DEST_PATH_IMAGE026
expressing the temperature of the intelligent park at any date after normalization processing;
Figure 419590DEST_PATH_IMAGE027
intelligent display gardenThe minimum temperature in the historical parking data within the zone,
Figure 833253DEST_PATH_IMAGE028
representing a maximum temperature in historical parking data for the smart park;
taking the intelligent park temperature time sequence data after normalization processing as a temperature influence factor after preprocessing;
and (3) forming time sequence data by the daily maximum parking quantity of the parking park k, and taking the formed time sequence data of the single-day maximum parking quantity as a parking quantity influence factor.
4. The intelligent park parking decision optimizing method of claim 1 wherein the step of S2 constructing a multiple influence factor intelligent park parking number regression model comprises:
for any parking lot k in the intelligent park, constructing a multi-influence-factor intelligent park parking quantity regression model, wherein the model takes the preprocessed parking influence factors as input and takes the prediction result of the parking quantity on the current date as output;
the regression model of the parking quantity of the multi-influence factor intelligent park is as follows:
Figure 451317DEST_PATH_IMAGE029
Figure 534810DEST_PATH_IMAGE030
Figure 503903DEST_PATH_IMAGE031
wherein:
Figure 455679DEST_PATH_IMAGE032
indicating stopThe parking number of the park k on the current date;
Figure 318461DEST_PATH_IMAGE033
representing a white noise sequence;
Figure 697490DEST_PATH_IMAGE034
represents the maximum number of parking lots k on a single day on the ith past day;
Figure 888300DEST_PATH_IMAGE035
is a regression coefficient, wherein
Figure 519133DEST_PATH_IMAGE036
；
c represents the number of parameters in the model, and L represents the time sequence length of the historical parking data of the parking park k;
substituting historical parking data of the parking lot k into a multiple influence factor intelligent lot parking quantity regression model, and solving by using a least square method to obtain a regression coefficient, wherein the solving process of the regression coefficient is as follows:
s21: respectively constructing the regression coefficients and the parking influence factors of the parking lot k into matrix forms, wherein the matrix form of the regression coefficients is
Figure 111788DEST_PATH_IMAGE037
T represents transposition;
the matrix form of the parking impact factors for parking park k is:
Figure 396139DEST_PATH_IMAGE038
Figure 957134DEST_PATH_IMAGE039
wherein:
Figure 250712DEST_PATH_IMAGE040
a parking impact factor representing the L th day in the past of the parking lot k;
Figure 838819DEST_PATH_IMAGE041
representing the historical single-day total parking of the park k,
Figure 559650DEST_PATH_IMAGE042
represents the total amount of single-day parking of the parking park k on the last L-th day;
s22: constructing a loss function of a regression model based on a least square method:
Figure 459473DEST_PATH_IMAGE043
wherein:
Figure 291163DEST_PATH_IMAGE044
a trace representing a computational matrix;
s23: calculating partial derivatives of the regression coefficients based on a loss function:
Figure 248624DEST_PATH_IMAGE045
let the left expression be 0, the solution result of the regression coefficient is:
Figure 140356DEST_PATH_IMAGE046
。
5. the intelligent park parking decision optimizing method of claim 4 wherein the step S3 is implemented by predicting the parking quantity of the parking park on the current date based on a multiple influence factor intelligent park parking quantity regression model and constructing the parking park parking feature vector, comprising:
based on the parking quantity of many influence factors wisdom garden parking quantity regression model prediction parking garden in the parking quantity of current date, then to the different parking garages in the wisdom garden, the current date parking quantity collection that the prediction obtained is:
Figure 668421DEST_PATH_IMAGE047
wherein:
Figure 303801DEST_PATH_IMAGE048
the parking number of the parking park K on the current date t is represented, and the K represents the total number of the parking parks in the intelligent park;
and (3) constructing parking characteristic vectors of the parking parks, wherein the parking characteristic vectors of any parking park k are as follows:
Figure 725555DEST_PATH_IMAGE049
wherein:
Figure 679867DEST_PATH_IMAGE050
indicating the location coordinates of the parking park k on the smart park,
Figure 554282DEST_PATH_IMAGE051
；
Figure 727775DEST_PATH_IMAGE052
representing the number of parking spaces in the parking park k;
Figure 144981DEST_PATH_IMAGE053
the result of the normalization process representing the hourly parking charge for the parking park k,
Figure 378516DEST_PATH_IMAGE054
，
Figure 740227DEST_PATH_IMAGE055
represents the maximum hourly parking charge for the intelligent campus,
Figure 107623DEST_PATH_IMAGE056
represents the minimum hourly parking charge for the intelligent campus,
Figure 238390DEST_PATH_IMAGE057
indicating an hourly parking charge for parking park k.
6. The intelligent park parking decision optimizing method according to claim 1, wherein the step of S4 constructing a neural network park decision optimizing model based on the park parking feature vectors comprises:
constructing a neural network parking park decision optimization model based on parking park parking feature vectors, wherein the decision optimization model takes a vehicle position and the parking feature vectors as input and takes an optimal parking park in an intelligent park as output;
the neural network parking area decision optimization model comprises an input layer, a convolution layer and a full connection layer, wherein the input of the input layer is vehicle position coordinates and parking characteristic vectors of different parking areas, and the input represents
Figure 642827DEST_PATH_IMAGE058
Comprises the following steps:
Figure 632780DEST_PATH_IMAGE059
wherein:
Figure 413654DEST_PATH_IMAGE060
as the position coordinates of the vehicle to be parked,
Figure 398927DEST_PATH_IMAGE061
；
Figure 863013DEST_PATH_IMAGE062
the parking characteristic vector of a parking park k in the intelligent park;
the input layer reconstructs the input representation and coordinates the position
Figure 199317DEST_PATH_IMAGE063
Distance of path from parking park position coordinate
Figure 783882DEST_PATH_IMAGE064
Adding the parking feature vector into the parking feature vector of the corresponding parking park, then:
Figure 764607DEST_PATH_IMAGE065
wherein:
Figure 510846DEST_PATH_IMAGE066
the updated parking characteristic vector of the parking park k is obtained;
the output representation of the input layer
Figure 600025DEST_PATH_IMAGE067
Comprises the following steps:
Figure 581757DEST_PATH_IMAGE068
to output the representation
Figure 541622DEST_PATH_IMAGE067
As input for convolutional layers, convolutional layer pairs
Figure 989921DEST_PATH_IMAGE067
And carrying out convolution processing to obtain a characteristic map representation of parking park characteristic vectors in the intelligent park, inputting the characteristic map representation into a full connection layer, outputting the probability of different parking parks as the optimal parking park by utilizing a softmax function in the full connection layer, selecting the parking with the maximum probability as the optimal parking park, and outputting the position coordinate of the optimal parking park.
7. The intelligent park parking decision optimizing method of claim 1 wherein the step of S5 inputting the vehicle position into the training optimized model, the model outputting the optimal parking park and sending the optimal parking park position to the vehicle navigation system comprises:
the position coordinates of the vehicle to be parked are input into the neural network parking park decision optimization model after training optimization, the model outputs the optimal parking park and sends the optimal parking park position to the vehicle navigation system, and a driver can drive the vehicle to arrive at the optimal parking park according to navigation information to park.
8. A smart park parking decision optimization system, the system comprising:
the parking quantity prediction module is used for acquiring parking influence factors of historical parking data in the intelligent park, preprocessing the acquired parking influence factors and constructing a multi-influence factor intelligent park parking quantity regression model;
the characteristic vector extraction device is used for constructing a parking characteristic vector of the parking park;
the parking decision module is used for constructing a neural network parking park decision optimization model based on parking park parking feature vectors, training and optimizing the neural network parking park decision optimization model by utilizing a heuristic algorithm, inputting the position of a vehicle into the model after training and optimization, outputting an optimal parking park by the model, and sending the optimal parking park position to a vehicle navigation system so as to realize the intelligent park parking decision optimization method according to any one of claims 1 to 7.
CN202210908844.0A 2022-07-29 2022-07-29 Intelligent park parking decision optimization method and system Active CN115049155B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210908844.0A CN115049155B (en) 2022-07-29 2022-07-29 Intelligent park parking decision optimization method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210908844.0A CN115049155B (en) 2022-07-29 2022-07-29 Intelligent park parking decision optimization method and system

Publications (2)

Publication Number Publication Date
CN115049155A true CN115049155A (en) 2022-09-13
CN115049155B CN115049155B (en) 2023-05-09

Family

ID=83166755

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210908844.0A Active CN115049155B (en) 2022-07-29 2022-07-29 Intelligent park parking decision optimization method and system

Country Status (1)

Country Link
CN (1) CN115049155B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170129537A1 (en) * 2015-11-10 2017-05-11 Hyundai Motor Company Method and apparatus for remotely controlling vehicle parking
CN108133619A (en) * 2018-02-06 2018-06-08 百度在线网络技术（北京）有限公司 Method, apparatus, storage medium and the terminal device of Parking prediction
CN108263374A (en) * 2016-12-30 2018-07-10 现代自动车株式会社 Automatic parking system and automatic stop process
CN109559553A (en) * 2018-12-12 2019-04-02 重庆邮电大学 A kind of wisdom shutdown system and its parking management method
CN114742327A (en) * 2022-06-10 2022-07-12 湖南前行科创有限公司 Rapid emergency disposal method and device for smart park based on collaborative optimization
CN114793182A (en) * 2022-06-21 2022-07-26 湖南前行科创有限公司 Intelligent park distributed network security risk assessment method and device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170129537A1 (en) * 2015-11-10 2017-05-11 Hyundai Motor Company Method and apparatus for remotely controlling vehicle parking
CN108263374A (en) * 2016-12-30 2018-07-10 现代自动车株式会社 Automatic parking system and automatic stop process
CN108133619A (en) * 2018-02-06 2018-06-08 百度在线网络技术（北京）有限公司 Method, apparatus, storage medium and the terminal device of Parking prediction
CN109559553A (en) * 2018-12-12 2019-04-02 重庆邮电大学 A kind of wisdom shutdown system and its parking management method
CN114742327A (en) * 2022-06-10 2022-07-12 湖南前行科创有限公司 Rapid emergency disposal method and device for smart park based on collaborative optimization
CN114793182A (en) * 2022-06-21 2022-07-26 湖南前行科创有限公司 Intelligent park distributed network security risk assessment method and device

Also Published As

Publication number Publication date
CN115049155B (en) 2023-05-09

Similar Documents

Publication Publication Date Title
WO2017063356A1 (en) Designated-driving order predicting method and designated-driving transport capacity scheduling method
CN107919014B (en) Taxi running route optimization method for multiple passenger mileage
CN111754025B (en) CNN+GRU-based short-time bus passenger flow prediction method
CN113780808B (en) Vehicle service attribute decision optimization method based on flexible bus connection system line
Kim et al. Idle vehicle relocation strategy through deep learning for shared autonomous electric vehicle system optimization
CN114119159A (en) Network taxi appointment real-time order matching and idle vehicle scheduling method and system
CN111340318B (en) Vehicle dynamic scheduling method and device and terminal equipment
CN108764579A (en) A kind of storage multi-robotic task dispatching method based on congestion control
CN112749653A (en) Pedestrian detection method, device, electronic equipment and storage medium
CN115730790A (en) Charging configuration method, device and equipment based on edge calculation and storage medium
CN109685411A (en) A kind of share-car matching process of logistics shipping platform
CN117252318B (en) Intelligent networking automobile group machine collaborative carpooling scheduling method and system
CN111126799A (en) Shared network driver and crew matching method based on bipartite graph
CN115130779A (en) Intelligent scheduling method based on mobile charging pile
CN114418606B (en) Network vehicle order demand prediction method based on space-time convolution network
CN115049155A (en) Intelligent park parking decision optimization method and system
CN114626766B (en) Shared electric vehicle scheduling method, device, equipment and medium based on big data
CN115131983A (en) Parking guidance method based on parking influence factors
CN113393111B (en) Cross-border transportation double-side connection vehicle scheduling method based on variable neighborhood tabu search algorithm
CN114971368A (en) Intelligent charging management and control method and system
CN112613802A (en) Parking space layout information generation method and three-dimensional parking system
CN114117883A (en) Self-adaptive rail transit scheduling method, system and terminal based on reinforcement learning
CN114202290A (en) Driving route recommendation method and device, electronic equipment and storage medium
CN114973745A (en) Parking lot recommendation method and automobile
WO2019128477A1 (en) Systems and methods for assigning service requests

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant