CN109886459A - A neural network-based parking demand forecasting method for public parking lots - Google Patents
A neural network-based parking demand forecasting method for public parking lots Download PDFInfo
- Publication number
- CN109886459A CN109886459A CN201910038428.8A CN201910038428A CN109886459A CN 109886459 A CN109886459 A CN 109886459A CN 201910038428 A CN201910038428 A CN 201910038428A CN 109886459 A CN109886459 A CN 109886459A
- Authority
- CN
- China
- Prior art keywords
- parking
- neural network
- layer
- rate
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 36
- 238000013277 forecasting method Methods 0.000 title claims description 13
- 238000000034 method Methods 0.000 claims abstract description 17
- 230000007306 turnover Effects 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 22
- 230000006870 function Effects 0.000 claims description 17
- 210000002569 neuron Anatomy 0.000 claims description 16
- 238000012937 correction Methods 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims 1
- 210000005036 nerve Anatomy 0.000 claims 1
- 238000007726 management method Methods 0.000 description 5
- 238000013439 planning Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000001154 acute effect Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明涉及一种基于神经网络的公共停车场停车需求预测方法，包括：步骤S1：采集各样本时段的需求特征值和对应的各影响因素的数据，以及待预测时段的各影响因素的数据，其中，需求特征值包括停车场泊位平均周转率、平均停车时长和各时段平均车位占有率；步骤S2：归一化处理各影响因素的数据；步骤S3：将各样本时段的需求特征值和对应的各影响因素的归一化数据对预测用神经网络进行训练和验证；步骤S4：将待预测时段的各影响因素的归一化数据输入训练好的神经网络，得到待预测时段的需求特征值。与现有技术相比，本发明可用于不同区域、能够反映停车收费对停车需求影响的公共停车场停车需求的预测。
The invention relates to a method for predicting parking demand in public parking lots based on a neural network, comprising: step S1: collecting demand characteristic values of each sample period and data of corresponding influencing factors, as well as data of each influencing factor of the period to be predicted, Among them, the demand characteristic value includes the average parking lot turnover rate, the average parking time and the average parking space occupancy rate in each period; Step S2: normalize the data of each influencing factor; Step S3: compare the demand characteristic value of each sample period with the corresponding The normalized data of each influencing factor is used to train and verify the neural network for prediction; Step S4: Input the normalized data of each influencing factor in the period to be predicted into the trained neural network to obtain the characteristic value of demand in the period to be predicted . Compared with the prior art, the present invention can be used for forecasting the parking demand of public parking lots in different areas, which can reflect the influence of parking charges on the parking demand.
Description
Technical field
The present invention relates to a kind of artificial intelligence technologys, stop more particularly, to a kind of Public Parking neural network based
Needing forecasting method.
Background technique
Parking lot is an extremely important ring in static traffic management, carries out the planning, operation, management of park construction,
The parking demand of vehicle can utmostly be met, reduce parking offense, promote road traffic efficiency., whereas if without root
It is averaged turnover rate, averagely stop duration, the parking characteristic customization pipe appropriate such as day part parking stall occupation rate according to the berth in parking lot
Reason strategy, on the one hand will cause parking difficulty problem and is difficult to be effectively relieved；On the other hand, if Public parking area is excessive or charged
Height leads to the waste of fund and land resource it will cause parking lot utilization efficiency is not high, influences society, expanding economy.Cause
This, Accurate Prediction Parking feature has emphatically parking lot of making rational planning for, formulation charge operation management strategy appropriate
Big meaning.
The prediction technique of parking demand has parking generation rate model, land used and Traffic Impact Analysis model, polynary time at present
Return analysis prediction model and the relevant parameter of model is planning, design with building parking demand standard etc., but in existing method
Personnel obtain from professional standard, specification or according to model recurrence, can not reflect to adaptation to local conditions different proposed regions, different land use
The complicated factors bring such as population, economy of property influences, and actual conditions often have deviation.Meanwhile the prior art can not
The accurate influence for considering Car park payment strategy for parking demand is unable to satisfy parking lot network operator and is formulating parking receipts appropriate
Take demand when strategy.
Such as Chinese patent CN 103093643A discloses a kind of method of determining Public Parking Berth number, utilizes
The method that BP neural network determines Public Parking Berth number.Its neural network input layer considered includes population, GDP, family
Front yard is averaged monthly income and family is averaged car number, output layer is the Berth number for planning parking lot.But this mode only can determine that
The berth number in parking lot, and it is unpredictable with the operation in built parking lot, the relevant berth of management, pricing strategy it is average
The characteristic parameters such as turnover rate, duration of averagely stopping, average parking stall occupation rate.
For another example Chinese patent CN 102867407A discloses a kind of parking lot effective berth occupation rate multistep forecasting method,
It utilizes wavelet neural network-maximum lyapunov exponent method combination forecasting, to parking lot effective berth occupation rate
Carry out multi-step prediction.But this mode can only carry out the prediction of berth occupation rate to a certain specific, built parking lot,
The parking feature in the parking lot of the different zones in planning can not be carried out pre- according to the Land character in the proposed region in parking lot
It surveys.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind to be based on neural network
Public Parking parking facilities' forecasting method.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of Public Parking parking facilities' forecasting method neural network based, comprising:
Step S1: the demand characteristic value of each sample period and the data of corresponding each influence factor and to be predicted are acquired
The data of each influence factor of period, wherein the demand characteristic value include parking lot berth be averaged turnover rate, averagely stop when
Long and day part is averaged parking stall occupation rate；
Step S2: the data of each influence factor of normalized；
Step S3: the normalization data of the demand characteristic value of each sample period and corresponding each influence factor uses prediction
Neural network is trained and verifies；
Step S4: the normalization data of each influence factor of period to be predicted is inputted into trained neural network, is obtained
The demand characteristic value of period to be predicted.
The influence factor includes 500 meters of periphery land-use style, 500 meters of periphery range job sum, 500 meters of periphery
Range residential home sum, 500 meters of periphery range family be averaged annual income, in 1000 range of periphery parking position it is total, to pre-
Parking lot is surveyed to park quantity, parking fee collective system standard to be predicted, park construction type to be predicted.
The neural network includes input layer, hidden layer and output layer, and the input layer includes 8 neurons, described defeated
Layer includes 3 neurons out, and the activation primitive of the hidden layer is Sigmoid function, and the neuron number of hidden layer are as follows:
Wherein: m is input layer number, and n is output layer neuron number, and α is constant, value range 1-10.
The Sigmoid function is smooth curve function, mathematic(al) representation are as follows:
Wherein: f (x) is the output of implicit layer unit, and x is the input of implicit layer unit.
The training process of neural network specifically includes in the step S3:
Step S31: initialization connection weight；
Step S32: choosing one group of sample from training set, using the demand characteristic value of the sample period as target sample,
Using the normalization data of corresponding each influence factor as input sample, it is supplied to neural network；
Step S33: with the input of input sample, connection weight and its threshold calculations hidden layer each unit, then pass through
The output of Sigmoid function calculating hidden layer each unit；
Step S34: it is missed using the response computation output layer each unit generalization of neural network object vector and each neuron
Difference；
Step S35: hidden layer each unit is calculated using the output of connection weight, the generalized error of output layer and hidden layer
Generalized error；
Step S36: adaptive regularized learning algorithm rate；
Step S37: each connection weight and its threshold value are corrected based on adaptive regularized learning algorithm rate；
Step S38: randomly selecting next group of training sample vector and be supplied to network, returns to step S33, until network is complete
Office's error function E is less than preset value, or if the number of iterations N reaches the upper limit；
Step S39: exporting the connection weight of each memory, and training terminates.
The step S36 is specifically included:
Step S361: the correction value of connection weight is calculated:
Wherein: Δ W (N) is the correction value of connection weight when iv-th iteration calculates, EpFor the side of actual value and desired output
Difference, W are the correction value of connection weight, and η is factor of momentum, and α is learning rate, and N is current iteration number；
Step S362: learning rate is calculated:
α (N)=α (N-1)+Δ α (N)
Wherein: Δ α (N) is the incrementss of learning rate when iv-th iteration learns, and α (N) is when iv-th iteration learns
Learning rate, a are the learning rate incrementss upper limit, and b is that learning rate increases coefficient of discharge, when α (N-1) is the N-1 times iterative learning
Habit rate, E (N) are training error.
Compared with prior art, the invention has the following advantages:
1) it can be used for different zones, be able to reflect the Public Parking parking demand that Car park payment influences parking demand
Prediction.
2) eight kinds of influence factors are introduced, and devise 8 corresponding input layers, improve the sophistication of data.
3) momentum term and automatic adjusument learning rate are increased to improve to algorithm, accelerates convergence process, improves and learns
Practise speed.
Detailed description of the invention
Fig. 1 is the key step flow diagram of the method for the present invention；
Fig. 2 is the structure chart of neural network of the present invention；
Fig. 3 is the flow chart of present invention training neural network.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
A kind of Public Parking parking facilities' forecasting method neural network based, as shown in Figure 1, comprising:
Step S1: the demand characteristic value of each sample period and the data of corresponding each influence factor and to be predicted are acquired
The data of each influence factor of period, wherein demand characteristic value include parking lot berth be averaged turnover rate, averagely stop duration and
Day part is averaged parking stall occupation rate；
The required data for obtaining influence factor include that 500 meters of periphery land-use style, 500 meters of periphery range job are total
Parking pool in number, 500 meters of periphery range residential home are total, 500 meters of periphery range family is averaged annual income, 1000 range of periphery
Position is total, parking lot to be predicted is parked quantity, parking fee collective system standard to be predicted, park construction type to be predicted；In acquisition
The data of factor are stated, and split data into training data group and prediction data group.
Wherein, 500 meters of periphery land-use style with a size be 1 × 5 land-use style matrix [a1, a2, a3, a4,
A5] it indicates, wherein a1-a5 respectively indicates business management shop (not including dining room) quantity within the scope of 500 meters of periphery, dining room number
Amount, school's number, hospital's number, movie theatre number.
Wherein, the parking fee collective system standard that need to be predicted with a size be 1 × 6 Car park payment matrix [b1, b2,
B3, b4, b5, b6] it indicates, wherein b1-b6 respectively indicates the parking rate that the parking lot predicted on demand is currently set, vehicle
Toll amount at the Parking 30 minutes, 2 hours, 4 hours, 6 hours, 8 hours, 10 hours.
Wherein, the construction type in the parking lot of need prediction is self-propelled Surface parking lots, self-propelled underground parking
Library, full-automatic mechanical parking building, semi automatic machine parking building, self-propelled spatial parking building, this one of six kinds of trackside berth.
Wherein, the mode for obtaining above-mentioned data includes on-site inspection, web crawlers extraction POI data, consults related official
Statistical yearbook etc..
It is corresponding, as shown in Fig. 2, the improved BP neural network used in the application includes input layer, hidden layer and defeated
Layer out, input layer include 8 neurons, and output layer includes 3 neurons, and the activation primitive of hidden layer is Sigmoid function, and
The neuron number of hidden layer are as follows:
Wherein: m is input layer number, and n is output layer neuron number, and α is constant, value range 1-10,
It is specifically determined by network training.
Sigmoid function is smooth curve function, mathematic(al) representation are as follows:
Wherein: f (x) is the output of implicit layer unit, and x is the input of implicit layer unit.
Step S2: the data of each influence factor of normalized, the process can use the life of the mapminmax in Matlab
It enables and realizing；
Step S3: the normalization data of the demand characteristic value of each sample period and corresponding each influence factor uses prediction
Neural network is trained and verifies, as shown in figure 3, wherein the training process of neural network specifically includes:
Step S31: initialization connection weight；
Step S32: choosing one group of sample from training set, using the demand characteristic value of the sample period as target sample,
Using the normalization data of corresponding each influence factor as input sample, it is supplied to neural network；
Step S33: with the input of input sample, connection weight and its threshold calculations hidden layer each unit, then pass through
The output of Sigmoid function calculating hidden layer each unit；
Step S34: it is missed using the response computation output layer each unit generalization of neural network object vector and each neuron
Difference；
Step S35: hidden layer each unit is calculated using the output of connection weight, the generalized error of output layer and hidden layer
Generalized error；
Step S36: adaptive regularized learning algorithm rate, i.e. improved BP and the improvements of former BP neural network exist
In increasing momentum term and automatic adjusument learning rate to improve to algorithm, accelerate convergence process, improve pace of learning.
Wherein, the stability of algorithm can be increased using momentum arithmetic with the acute variation in smooth gradient direction, accelerates net
The convergence rate of network.After increasing momentum term, modified weight amount added the memory in relation to last moment weight modification direction, that is, work as
Between the moment modification direction be last moment modify direction and current time direction combination,
It is specific:
Step S361: the correction value of connection weight is calculated:
Wherein: Δ W (N) is the correction value of connection weight when iv-th iteration calculates, EpFor the side of actual value and desired output
Difference, W are correction value, and η is factor of momentum, and α is learning rate, and N is current iteration number；
Step S362: automatic adjusument learning rate refers in the training process, when iv-th iteration learns, if first N-1 times changes
Learning rate α (N-1) when generation study unanimously tends to training error E (N) decline, then α (N) is intended to increase；Conversely, if consistent
It must tend to increase training error E (N), then α (N) tends to reduce；When training error is constant, learning rate is remained unchanged,
Specifically, calculating learning rate:
α (N)=α (N-1)+Δ α (N)
Wherein: Δ α (N) is the incrementss of learning rate when iv-th iteration learns, and α (N) is when iv-th iteration learns
Learning rate, a are the learning rate incrementss upper limit, and b is that learning rate increases coefficient of discharge, when α (N-1) is the N-1 times iterative learning
Habit rate, E (N) are training error.
Step S37: each connection weight and its threshold value are corrected based on adaptive regularized learning algorithm rate；
Step S38: randomly selecting next group of training sample vector and be supplied to network, returns to step S33, until network is complete
Office's error function E is less than preset value, or if the number of iterations N reaches the upper limit；
Step S39: exporting the connection weight of each memory, and training terminates.
Using gradient descent method adjustment hidden layer to input layer and input layer to the weight of hidden layer, pass through Matlab tool
Traingdm order is realized in case.
Wherein, allowable error range value set by the error can be ± 5%, ± 10%, ± 15%, ± 20%
It is equivalent.The calculation method of error are as follows:
Wherein
Parking demand characteristic value refers to that the berth in the parking lot is averaged turnover rate, averagely stop duration and day part parking stall average occupancy
3 characteristic values.
Wherein, day part parking stall average occupancy is by (06:00-11:00), noon (11:00-16:00), night between morning
(17:00-22:00), the late into the night (22:00-06:00) are divided into four periods.
Step S4: being input to trained improved BP for the data that prediction data is concentrated, every in step S1
A influence factor for influencing Public Parking parking demand corresponds to 1 neuron of improved BP input layer, input layer
Data the data of hidden layer are generated by the processing of S type function, imply layer data and generated by the processing of linear transfer function
The data of output layer, i.e. berth be averaged turnover rate, averagely stop 3 duration, day part parking stall average occupancy data.
Claims (6)
1. a kind of Public Parking parking facilities' forecasting method neural network based characterized by comprising
Step S1: the demand characteristic value of each sample period and data and the period to be predicted of corresponding each influence factor are acquired
Each influence factor data, wherein the demand characteristic value include parking lot berth be averaged turnover rate, averagely stop duration and
Day part is averaged parking stall occupation rate；
Step S2: the data of each influence factor of normalized；
Step S3: by the normalization data of the demand characteristic value of each sample period and corresponding each influence factor to prediction nerve
Network is trained and verifies；
Step S4: the normalization data of each influence factor of period to be predicted is inputted into trained neural network, is obtained to pre-
Survey the demand characteristic value of period.
2. a kind of Public Parking parking facilities' forecasting method neural network based according to claim 1, feature
It is, the influence factor includes 500 meters of periphery land-use style, 500 meters of periphery range job sum, 500 meters of periphery model
Enclose residential home sum, 500 meters of periphery range family be averaged annual income, parking position is total, to be predicted in 1000 range of periphery
Parking lot is parked quantity, parking fee collective system standard to be predicted, park construction type to be predicted.
3. a kind of Public Parking parking facilities' forecasting method neural network based according to claim 2, feature
It is, the neural network includes input layer, hidden layer and output layer, and the input layer includes 8 neurons, the output layer
Including 3 neurons, the activation primitive of the hidden layer is Sigmoid function, and the neuron number of hidden layer are as follows:
Wherein: m is input layer number, and n is output layer neuron number, and α is constant, value range 1-10.
4. a kind of Public Parking parking facilities' forecasting method neural network based according to claim 3, feature
It is, the Sigmoid function is smooth curve function, mathematic(al) representation are as follows:
Wherein: f (x) is the output of implicit layer unit, and x is the input of implicit layer unit.
5. a kind of Public Parking parking facilities' forecasting method neural network based according to claim 3, feature
It is, the training process of neural network specifically includes in the step S3:
Step S31: initialization connection weight；
Step S32: choosing one group of sample from training set, will be right using the demand characteristic value of the sample period as target sample
The normalization data for each influence factor answered is supplied to neural network as input sample；
Step S33: with the input of input sample, connection weight and its threshold calculations hidden layer each unit, then pass through Sigmoid
The output of function calculating hidden layer each unit；
Step S34: the response computation output layer each unit generalization error of neural network object vector and each neuron is utilized；
Step S35: the one of hidden layer each unit is calculated using the output of connection weight, the generalized error of output layer and hidden layer
As change error；
Step S36: adaptive regularized learning algorithm rate；
Step S37: each connection weight and its threshold value are corrected based on adaptive regularized learning algorithm rate；
Step S38: randomly selecting next group of training sample vector and be supplied to network, returns to step S33, until the network overall situation is missed
Difference function E is less than preset value, or if the number of iterations N reaches the upper limit；
Step S39: exporting the connection weight of each memory, and training terminates.
6. a kind of Public Parking parking facilities' forecasting method neural network based according to claim 5, feature
It is, the step S36 is specifically included:
Step S361: the correction value of connection weight is calculated:
Wherein: Δ W (N) is the correction value of connection weight when iv-th iteration calculates, EpFor the variance of actual value and desired output, W
For the correction value of connection weight, η is factor of momentum, and α is learning rate, and N is current iteration number；
Step S362: learning rate is calculated:
α (N)=α (N-1)+Δ α (N)
Wherein: Δ α (N) is the incrementss of learning rate when iv-th iteration learns, and α (N) is study when iv-th iteration learns
Rate, a are the learning rate incrementss upper limit, and b increases coefficient of discharge for learning rate, learning rate when α (N-1) is the N-1 times iterative learning,
E (N) is training error.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910038428.8A CN109886459A (en) | 2019-01-16 | 2019-01-16 | A neural network-based parking demand forecasting method for public parking lots |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910038428.8A CN109886459A (en) | 2019-01-16 | 2019-01-16 | A neural network-based parking demand forecasting method for public parking lots |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109886459A true CN109886459A (en) | 2019-06-14 |
Family
ID=66926069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910038428.8A Pending CN109886459A (en) | 2019-01-16 | 2019-01-16 | A neural network-based parking demand forecasting method for public parking lots |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109886459A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110503104A (en) * | 2019-08-29 | 2019-11-26 | 桂林电子科技大学 | A short-term prediction method for the number of remaining parking spaces based on convolutional neural network |
| CN110599236A (en) * | 2019-08-14 | 2019-12-20 | 同济大学 | Short-time parking demand prediction method based on GRU model |
| CN110659774A (en) * | 2019-09-23 | 2020-01-07 | 北京交通大学 | Parking demand forecasting method driven by big data method |
| CN110992700A (en) * | 2019-10-31 | 2020-04-10 | 江苏新弈软件有限公司 | Parking data acquisition method based on automatic detection technology |
| CN113326962A (en) * | 2020-02-28 | 2021-08-31 | 南京理工大学 | Subway train ATO speed curve prediction method based on BP neural network |
| CN113380071A (en) * | 2021-06-25 | 2021-09-10 | 南京静态交通产业技术研究院 | Parking space level parking demand state evaluation method for large complex |
| CN113705891A (en) * | 2021-08-27 | 2021-11-26 | 南京林业大学 | Urban business office complex building parking demand prediction method based on MRA-BAS-BP algorithm |
| CN114048886A (en) * | 2021-09-15 | 2022-02-15 | 深圳大学 | Effective berth combination prediction method, terminal and storage medium based on neural network |
| CN119229682A (en) * | 2024-12-05 | 2024-12-31 | 山东通维信息工程有限公司 | Smart parking lot management method and cloud platform based on information sharing |
| CN119989956A (en) * | 2025-04-17 | 2025-05-13 | 东华理工大学南昌校区 | A temperature prediction method and system for temperature derivatives |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105956690A (en) * | 2016-04-25 | 2016-09-21 | 广州东芝白云自动化系统有限公司 | Water supply prediction method and water supply prediction system |
| CN106815782A (en) * | 2017-01-23 | 2017-06-09 | 重庆汇集源科技有限公司 | A kind of real estate estimation method and system based on neutral net statistical models |
| KR20180000625A (en) * | 2016-06-23 | 2018-01-03 | 주식회사 케이티 | Apparatus for predicting diffusion of epidemic animal disease and method thereof |
| CN108417032A (en) * | 2018-03-19 | 2018-08-17 | 中景博道城市规划发展有限公司 | A Method for Analysis and Prediction of On-street Parking Demand in Urban Central Area |
| CN108961816A (en) * | 2018-07-19 | 2018-12-07 | 泰华智慧产业集团股份有限公司 | Road parking berth prediction technique based on optimization LSTM model |
| CN109215380A (en) * | 2018-10-17 | 2019-01-15 | 浙江科技学院 | A kind of prediction technique of effective parking position |
-
2019
- 2019-01-16 CN CN201910038428.8A patent/CN109886459A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105956690A (en) * | 2016-04-25 | 2016-09-21 | 广州东芝白云自动化系统有限公司 | Water supply prediction method and water supply prediction system |
| KR20180000625A (en) * | 2016-06-23 | 2018-01-03 | 주식회사 케이티 | Apparatus for predicting diffusion of epidemic animal disease and method thereof |
| CN106815782A (en) * | 2017-01-23 | 2017-06-09 | 重庆汇集源科技有限公司 | A kind of real estate estimation method and system based on neutral net statistical models |
| CN108417032A (en) * | 2018-03-19 | 2018-08-17 | 中景博道城市规划发展有限公司 | A Method for Analysis and Prediction of On-street Parking Demand in Urban Central Area |
| CN108961816A (en) * | 2018-07-19 | 2018-12-07 | 泰华智慧产业集团股份有限公司 | Road parking berth prediction technique based on optimization LSTM model |
| CN109215380A (en) * | 2018-10-17 | 2019-01-15 | 浙江科技学院 | A kind of prediction technique of effective parking position |
Non-Patent Citations (2)
| Title |
|---|
| 嵇春艳: "《海洋平台随机动力响应分析方法及智能控制技术》", 31 March 2013, 上海交通大学出版社 * |
| 张济民: "基于神经网络的预测控制在摆式客车倾摆系统的应用研究", 《中国优秀博硕士学位论文全文数据库 (博士) 工程科技Ⅱ辑》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110599236A (en) * | 2019-08-14 | 2019-12-20 | 同济大学 | Short-time parking demand prediction method based on GRU model |
| CN110503104A (en) * | 2019-08-29 | 2019-11-26 | 桂林电子科技大学 | A short-term prediction method for the number of remaining parking spaces based on convolutional neural network |
| CN110503104B (en) * | 2019-08-29 | 2021-04-27 | 桂林电子科技大学 | A Short-term Remaining Parking Space Prediction Method Based on Convolutional Neural Network |
| CN110659774B (en) * | 2019-09-23 | 2022-08-02 | 北京交通大学 | Parking demand forecasting method driven by big data method |
| CN110659774A (en) * | 2019-09-23 | 2020-01-07 | 北京交通大学 | Parking demand forecasting method driven by big data method |
| CN110992700A (en) * | 2019-10-31 | 2020-04-10 | 江苏新弈软件有限公司 | Parking data acquisition method based on automatic detection technology |
| CN113326962A (en) * | 2020-02-28 | 2021-08-31 | 南京理工大学 | Subway train ATO speed curve prediction method based on BP neural network |
| CN113380071A (en) * | 2021-06-25 | 2021-09-10 | 南京静态交通产业技术研究院 | Parking space level parking demand state evaluation method for large complex |
| CN113380071B (en) * | 2021-06-25 | 2022-09-06 | 南京静态交通产业技术研究院 | Parking space level parking demand state evaluation method for large complex |
| CN113705891A (en) * | 2021-08-27 | 2021-11-26 | 南京林业大学 | Urban business office complex building parking demand prediction method based on MRA-BAS-BP algorithm |
| CN113705891B (en) * | 2021-08-27 | 2024-02-13 | 南京林业大学 | Urban commercial complex building parking demand prediction method based on MRA-BAS-BP algorithm |
| CN114048886A (en) * | 2021-09-15 | 2022-02-15 | 深圳大学 | Effective berth combination prediction method, terminal and storage medium based on neural network |
| CN119229682A (en) * | 2024-12-05 | 2024-12-31 | 山东通维信息工程有限公司 | Smart parking lot management method and cloud platform based on information sharing |
| CN119229682B (en) * | 2024-12-05 | 2025-02-25 | 山东通维信息工程有限公司 | Intelligent parking lot management method based on information sharing and cloud platform |
| CN119989956A (en) * | 2025-04-17 | 2025-05-13 | 东华理工大学南昌校区 | A temperature prediction method and system for temperature derivatives |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109886459A (en) | A neural network-based parking demand forecasting method for public parking lots | |
| CN110288212B (en) | Site selection method for new electric taxi charging station based on improved MOPSO | |
| CN110555990B (en) | Effective parking space-time resource prediction method based on LSTM neural network | |
| CN103177570B (en) | Method for predicting traffic jam indexes for rush hours in morning and evening | |
| CN110299008B (en) | Traffic flow multi-step prediction method based on reinforcement learning | |
| CN107886186A (en) | A kind of charging pile method to set up based on travelling data and Wei Nuotu zonings | |
| CN112907103B (en) | A method for dynamic supply and demand balance of shared bicycles | |
| CN108417032A (en) | A Method for Analysis and Prediction of On-street Parking Demand in Urban Central Area | |
| CN113705922B (en) | An improved ultra-short-term wind power power prediction algorithm and model establishment method | |
| CN110309436B (en) | Automobile service network collaborative site selection method and system considering owner selection behavior | |
| CN112613637B (en) | Method and device for processing charging load | |
| CN107122856A (en) | Space saturation load forecasting method under new situation | |
| CN105761192A (en) | Intelligent method and intelligent integrated system for village-town area land use planning | |
| CN112419711B (en) | Prediction method of parking demand in closed parking lot based on improved GMDH algorithm | |
| CN118968814B (en) | Intelligent parking charging pile positioning method and system combined with solar energy | |
| Schall | Urban renewal policy and economic efficiency | |
| CN112906945A (en) | Traffic flow prediction method, system and computer readable storage medium | |
| CN114811713A (en) | Secondary inter-user balanced heat supply regulation and control method based on hybrid deep learning | |
| CN118968790A (en) | Multi-traffic signal light control method and system based on multi-agent reinforcement learning | |
| Zhao et al. | Forecasting regional short-term freight volume using QPSO-LSTM algorithm from the perspective of the importance of spatial information | |
| Kepaptsoglou et al. | Optimizing pricing policies in Park-and-Ride facilities: a model and decision support system with application | |
| CN114298379A (en) | An optimization method for automatic valet parking lot layout based on simulation optimization | |
| CN119151618A (en) | Network vehicle cost scheduling method based on supply and demand prediction and storage medium | |
| CN117556186A (en) | Urban bus travel demand mode estimation method driven by multi-source data | |
| CN117787585A (en) | Method and device for predicting number of charging piles driven by data |
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: 20190614 |