CN111159533B - Intelligent charging service recommendation method and system based on user image - Google Patents
Intelligent charging service recommendation method and system based on user image Download PDFInfo
- Publication number
- CN111159533B CN111159533B CN201911203017.6A CN201911203017A CN111159533B CN 111159533 B CN111159533 B CN 111159533B CN 201911203017 A CN201911203017 A CN 201911203017A CN 111159533 B CN111159533 B CN 111159533B
- Authority
- CN
- China
- Prior art keywords
- charging station
- user
- charging
- initial
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000013528 artificial neural network Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims description 60
- 238000004364 calculation method Methods 0.000 claims description 31
- 238000012545 processing Methods 0.000 claims description 17
- 238000001914 filtration Methods 0.000 claims description 13
- 230000007613 environmental effect Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 11
- XOFYZVNMUHMLCC-ZPOLXVRWSA-N prednisone Chemical compound O=C1C=C[C@]2(C)[C@H]3C(=O)C[C@](C)([C@@](CC4)(O)C(=O)CO)[C@@H]4[C@@H]3CCC2=C1 XOFYZVNMUHMLCC-ZPOLXVRWSA-N 0.000 claims description 8
- 238000003062 neural network model Methods 0.000 claims description 6
- 230000036541 health Effects 0.000 claims description 3
- 230000036962 time dependent Effects 0.000 claims description 2
- 230000006399 behavior Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 7
- 230000005611 electricity Effects 0.000 description 6
- 238000009826 distribution Methods 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Economics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Public Health (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The application provides an intelligent charging service recommending method and system based on user image, comprising the following steps: obtaining an initial selectable charging station based on the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources; determining a pushable charging station from the initial selectable charging stations based on pre-obtained user portrait tag weights; the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data. By means of the scheme, the charging demand under the user travel scene determined by the current electric automobile state and the user portrait can be accurately predicted, and then an optimal charging station selection scheme is provided for the user.
Description
Technical Field
The application relates to the field of electric vehicle charging, in particular to an intelligent recommendation system of an electric vehicle charging station.
Background
The recommendation system is a personalized information push system for recommending information, products and the like which are interested by the user to the user according to the information requirements, interests and the like of the user. A good recommendation system not only can provide personalized service for users, but also can establish close relation with users, so that the users can rely on recommendation. In recent years, the development of fields such as machine learning, deep learning and the like provides method guidance for a recommendation system, so that the recommendation system is widely applied to a plurality of fields, wherein the most typical field with good development and application prospects is the electronic commerce field.
However, in the field of charging service in the electric car post-service market, the application of the recommendation system is not mature. The reason for this is the specificity of the recommended "commodity" in the context of the charging service. The "commodity" is essentially electric energy, similar to a gas station, but when the user is provided with energy services, factors influencing the decision of the user often exceed the properties of the commodity itself, such as site location, traffic conditions, etc. Meanwhile, different from gasoline, the attribute of the electric energy provided by the charging station is also complex, including: 1) The quality of the fuel is similar to that of the gasoline, the service life of the gasoline pump is influenced, and the service life of the power battery is also influenced by the charging power; 2) Because the charging time is long and the number of charging stations is small, the queuing time of the charging station is far longer than that of the gas station; 3) Different from the price of refueling according to the area, the charging cost is mainly determined by the electricity price, the charging service charge and the parking charge, the electricity price is generally peak-valley electricity price at present, the electricity price has larger fluctuation at different moments in the day, and the randomness of the charging charge is increased along with the introduction of a dynamic pricing mechanism of the charging service charge; 4) Although there is no difference in the quality of the electrical energy at the demand side, the electrical energy supplied by the user may come from a conventional thermal power plant or renewable energy source at the supply side, and as the carbon emission rights are tried in some cities, whether the charged energy is "green electricity" is one of the potential influencing factors that may influence the decision of the user in the future. In summary, the "commodity" attribute of the charging service has great randomness and uncertainty, so how to improve the matching between the charging recommendation scheme and the real requirement of the user in the charging station recommendation system is an important problem to be solved by those skilled in the art.
Disclosure of Invention
In order to improve the matching property of a charging recommendation scheme in a charging station recommendation system and the real requirement of a user, the application adopts the following technical scheme:
an intelligent charging service recommending method based on user images comprises the following steps:
obtaining an initial selectable charging station based on the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources;
determining a pushable charging station from the initial selectable charging stations based on pre-obtained user portrait tag weights;
the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data.
Preferably, training the user portrayal tag weights includes:
performing label matching on the characteristic data in the user order data and the user behavior data based on the set user portrait label to generate a training log;
processing the training log by a characteristic hash and low-frequency filtering method or an equal-frequency discretizing method to obtain training data;
the training data are put into a multi-task deep neural network model for training, and user portrait tag weights are obtained;
the user portrait tag includes: open time, idle rate, distance, charge power, cost, and environment;
the feature data includes: user characteristic data, asset characteristic data, and environmental characteristic data.
Preferably, the training log is processed by a feature hash and low-frequency filtering method, which comprises the following steps:
converting the training log into a real number matrix group through characteristic hash;
and performing low-frequency filtering processing on the discrete features of the real number matrix group, and removing features smaller than an occurrence frequency threshold value to form the training data.
Preferably, the training log is processed by an equal frequency discretization method, which comprises the following steps:
the data in the training log is divided according to fixed frequency to obtain several groups of equal sample sizes.
Preferably, based on the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources, obtaining the initial selectable charging station includes:
calculating the safe driving mileage of the electric automobile in the current state;
determining the safe driving range of the electric automobile by taking the current coordinate position of the electric automobile as a circle center and the safe driving mileage as a radius;
setting all charging stations with available states in the safe driving range of the electric automobile as initial selectable charging stations;
the remaining driving mileage of the electric vehicle in the current state is determined by the safety driving mileage.
Preferably, the calculation formula of the remaining driving mileage of the electric vehicle in the current state is as follows:
wherein S is rest For remaining mileage, SOC 1 And (3) the power battery charge quantity at the current coordinate position, SOH is the health state of the power battery, C is the rated capacity of the power battery, V is the current voltage of the power battery, and InitEC is the energy consumed by the unit driving mileage.
Preferably, determining a pushable charging station from the initial selectable charging stations based on pre-obtained user portrait tag weights comprises:
calculating a user charging position decision-making influence factor value of each power station in the initial selectable charging station;
determining a push charging station based on a pre-obtained user portrait tag weight and a user charging position decision influence factor value of each power station;
the corresponding label pushing client is marked on the pushing charging station;
the user charging position decision influence factor value of the power station comprises: time-impact factor value, cost-impact factor value, and environmental-impact factor value.
Preferably, the calculation formula for determining the push-able charging station is as follows:
Rank＝α·ΔT+β·P+γ·X
wherein Rank is an initial optional power station score, alpha is a time tag weight, beta is a cost tag weight, gamma is an environmental tag weight, deltaT is a time influence factor value, P is a cost influence factor value, and X is an environmental influence factor value.
Preferably, the time-dependent factor value is calculated as follows:
△T＝△T 2 +△T 3 +△T 4 +△T 5
wherein DeltaT is the time influence factor value DeltaT 2 For the travel time of the electric vehicle from the current position to the initial selectable charging station position, deltaT 3 For queuing time, deltaT after reaching an initially selectable charging station 4 Paying time for initial selectable charging station, deltaT 5 The electric vehicle is charged for a time at an initial optional charging station.
Preferably, the electric vehicle is charged for a time DeltaT at the initial selectable charging station 5 The calculation formula is as follows:
wherein delta E is the charging quantity of the electric automobile, and P Work (work) Is the charging power.
Preferably, the calculation formula of the charging electric quantity delta E of the electric automobile is as follows:
ΔE＝{SOC 2 -(SOC 1 +ΔSOC)}×C
wherein SOC is 2 Battery charge and SOC for electric vehicle 1 The delta SOC is the battery charge loss of the electric vehicle from the current position to the initial optional charging station positionAnd C is the rated capacity of the battery of the electric automobile.
Preferably, the cost impact factor value P is calculated as follows:
P＝[P 1 +P 2 ]×△E+P 3 ×(△T 3 +△T 4 +△T 5 )；
wherein P is 1 Charging unit price, P for initial optional charging station 2 Charging service charge for initial optional charging station, delta E is charging quantity of electric automobile and P 3 Parking fee for initial optional charging station, deltaT 3 For queuing time, deltaT after reaching an initially selectable charging station 4 Paying time for initial selectable charging station, deltaT 5 The electric vehicle is charged for a time at an initial optional charging station.
Preferably, the time stamp weight α is calculated as follows:
α＝α 1 +α 2 +α 3 +α 4
wherein alpha is 1 Is the weight of the open time tag, alpha 2 For idle rate tag weight, alpha 3 Is the distance label weight alpha 4 The charge power is weighted.
Based on the same inventive concept, the application also provides an intelligent charging service recommendation system based on the user image, which comprises the following steps:
an initial selectable charging station summoning module and a pushable charging station determining module;
the initial selectable charging station recall module is used for obtaining an initial selectable charging station according to the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources;
the pushed charging station determining module is used for determining the pushed charging station from the initial selectable charging stations according to the pre-obtained user portrait tag weight;
the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data.
Preferably, the intelligent charging service recommendation system based on user portraits further comprises a user portrayal tag weight training module, wherein the user portrayal tag weight training module comprises:
the device comprises a label matching unit, a data processing unit and a data training unit;
the label base matching unit is used for carrying out label matching on the characteristic data in the user order data and the user behavior data according to the set user portrait label so as to generate a training log;
the data processing unit is used for processing the training log through a characteristic hash and low-frequency filtering method or an equal-frequency discretizing method to obtain training data;
the data training unit is used for bringing the training data into a multi-task deep neural network model for training to obtain user portrait tag weights;
the user portrait tag includes: open time, idle rate, distance, charge power, cost, and environment;
the feature data includes: user characteristic data, asset characteristic data, and environmental characteristic data.
Preferably, the initial optional charging station summoning module comprises:
a safe driving range calculation unit, a safe driving range calculation unit and an initial selectable charging station determination unit;
the safe driving mileage calculation unit is used for calculating the safe driving mileage of the electric automobile in the current state;
the safe driving range calculation unit is used for determining the safe driving range of the electric automobile by taking the position of the electric automobile as a circle center and the safe driving mileage as a radius;
the initial selectable charging station determining unit is used for setting all charging stations with available states in the safe driving range as initial selectable charging stations;
the remaining driving mileage of the electric vehicle in the current state is determined by the safety driving mileage.
Preferably, the push-able charging station determination module comprises:
the mobile terminal comprises a user charging position decision influence factor value calculation unit, a pushable charging station determination unit and a pushing unit;
the user charging position decision influence factor value calculation unit is used for calculating user charging position decision influence factor values of all power stations in the initial selectable charging station;
the determination unit of the push-able charging station is used for determining the push-able charging station according to the pre-obtained user portrait tag weight and the user charging position decision influence factor value of each power station;
the pushing unit is used for pushing the corresponding label to the client on the pushing charging station;
the user charging position decision influence factor value of the power station comprises: time-impact factor value, cost-impact factor value, and environmental-impact factor value.
Compared with the closest prior art, the technical scheme provided by the application has the following beneficial effects:
the application provides an intelligent charging service recommending method and system based on user image, comprising the following steps: obtaining an initial selectable charging station based on the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources; determining a pushable charging station from the initial selectable charging stations based on pre-obtained user portrait tag weights; the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data. By means of the scheme, the charging demand under the user travel scene determined by the current electric automobile state and the user portrait can be accurately predicted, and then an optimal charging station selection scheme is provided for the user.
Drawings
FIG. 1 is a schematic flow chart of a method for recommending intelligent charging service based on user image;
FIG. 2 is a schematic diagram of a basic structure of an intelligent charging service recommendation system based on user images according to the present application;
FIG. 3 is a detailed schematic diagram of an intelligent charging service recommendation system based on user images according to the present application;
FIG. 4 is a schematic diagram of a technical route of an intelligent charging service recommendation system based on user images according to an embodiment of the present application;
fig. 5 is a schematic diagram of a product prototype of an intelligent charging service recommendation system based on user images according to an embodiment of the present application.
Detailed Description
The application is described in further detail below with reference to the accompanying drawings.
Example 1:
the recommendation system provided in this embodiment binds with a user to implement data collection and interaction, where the user may designate at least one electric automobile to be used, and the flow of the recommendation method is shown in fig. 1, and includes: obtaining an initial selectable charging station based on the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources; determining a pushable charging station from the initial selectable charging stations based on pre-obtained user portrait tag weights; the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data.
Training the feature data and the user behavior data in the user order data based on the multi-task deep neural network to generate user portrait tag weights, including:
determining a label system of the user portrait, which comprises open time, idle rate, distance, charging power and cost, according to a user charging position decision influence factor value received by the mobile APP when the user generates a charging demand;
based on user order data of the national power grid intelligent car networking platform, the flexibility coefficient of charging prices, the sensitivity of charging pile information, car pile distance, road conditions and other information of various users are analyzed, and training of 5 user portrait tag weights is completed, as shown in fig. 4.
The recommended flow includes training log generation, training data generation, model training, on-line scoring and other stages. When the recommendation system recommends the user generating the charging requirement, the user characteristic data, the asset characteristic data and the environment characteristic data at the moment are recorded, and the user behavior feedback of the recommendation is collected.
Tag matching: the recommendation background log records user characteristics, asset characteristics and environment characteristics corresponding to the current sample, and the tag log captures behavior feedback of the user on the recommendation item. The system splices the two data together according to the unique ID to generate an original training log.
The input requirement of most machine learning algorithms is a real matrix, so the process called feature hashing (compressing the original high-dimensional feature vector into a lower-dimensional feature vector without losing the expressive power of the original features as much as possible) needs to convert the original data into a real matrix first.
Too many very sparse discrete features can cause overfitting problems during training (making assumptions overly stringent for consistent assumptions), while increasing the number of parameters stored. To avoid this problem, low frequency filtering is performed on the discrete features, discarding features less than the frequency of occurrence threshold.
Through analysis of training data, the value distribution of the features in different dimensions and the difference of the feature values in the same dimension can be found to be large. The data of features such as distance, price and the like obey long tail distribution, and the feature values of most samples are smaller, and the feature values of a few samples are very large. Therefore, in practice, normalization is performed according to the position of the eigenvalue in the cumulative distribution function, that is, the eigenvalue is subjected to equal frequency division into bins, so as to ensure that the sample size in each bin is basically equal. The method ensures that the features of different distributions can be mapped to approximately uniform distribution, thereby ensuring the distinguishing degree of the features among samples and the stability of the numerical value.
The user tag weights are trained using a multi-tasking deep neural network model, even if the input layer has n_in neurons and the output layer has n_out neurons. Plus some hidden layers containing several neurons. At this time, it is necessary to find the linear coefficient matrix W and the bias vector b corresponding to all the hidden layers and the output layers, so that the output calculated by all the training sample inputs is equal to or very close to the sample output as possible.
When the user portrait tag weight matrix meeting the evaluation effect function is obtained, the system stores dynamic data as static data.
When a user generates a charging demand, an initial optional charging station is first determined based on current coordinate information of the electric vehicle:
according to the state of charge of the battery of the electric automobile, road conditions, environment, air conditioner use states and the like, the remaining driving mileage of the electric automobile is calculated, and the calculation formula of the remaining driving mileage is as follows:
wherein S is rest For remaining mileage, SOC 1 SOH is the state of health of the power battery, C is the rated capacity of the power battery, V is the current voltage of the power battery and InitEC is the energy consumed by the unit driving mileage;
determining the current safe driving mileage of the electric automobile according to 80% of the remaining driving mileage of the electric automobile;
drawing a circle by taking the current position of the electric automobile as a circle center and the safe driving mileage as a radius, and determining the safe driving range of the electric automobile;
searching all charging stations in the available state in the safe driving range, and returning to the recommendation system as initial selectable charging stations.
Secondly, sorting and pushing the initial selectable charging stations to the user according to the stored user portrait tag weight matrix, wherein the method comprises the following steps:
according to the user portrait tag weight, the environmental characteristics and the initial selectable charging station state at the current moment, ranking and scoring are carried out on all charging stations in the initial selectable charging stations, and the calculation formula is as follows:
Rank＝α·ΔT+β·P+γ·X
rank is the initial selectable plant score, α is the time tag weight, β is the cost tag weight, γ is the environmental tag weight, Δt is the time influencing factor value, P is the cost influencing factor value, and X is the environmental influencing factor value;
wherein, the calculation formula of the time influence factor value DeltaT is as follows:
△T＝△T 2 +△T 3 +△T 4 +△T 5
△T 2 for the travel time of the electric vehicle from the current position to the initial selectable charging station position, deltaT 3 For queuing time, deltaT after reaching an initially selectable charging station 4 Paying time for initial selectable charging station, deltaT 5 Charging time for the electric vehicle at an initial selectable charging station;
wherein, the calculation formula of the cost impact factor value P is as follows:
P＝[P 1 +P 2 ]×△E+P 3 ×(△T 3 +△T 4 +△T 5 )；
P 1 charging unit price, P for initial optional charging station 2 Charging service charge for initial optional charging station, delta E is charging quantity of electric automobile and P 3 Parking fee for initial optional charging station, deltaT 3 For queuing time, deltaT after reaching an initially selectable charging station 4 Paying time for initial selectable charging station, deltaT 5 Charging time for the electric vehicle at an initial selectable charging station;
the charging decision influence factor value is obtained in the following manner:
△T 2 ＝t 3 -t 2 : calling Gaoder API interface and inputting Loc 2 ，t 2 And Loc 3 Return t 3 ；
△T 3 : an access vehicle volume statistical model of each station needs to be obtained according to historical order data, for example: fitting to obtain the approximate Poisson distribution P (lambda) of the number of the electric vehicles charged to the charging station A on a certain working day, and further obtaining the charging station A according to the idle state and DeltaT of the current station 2 The number of arrival stations in between predicts the queuing time.
△T 4 : the code scanning payment response time is optimized by other teams and is set as in project period1s；
△T 5 Charging time DeltaT of electric automobile at initial optional charging station 5 The calculation formula is as follows:
delta E is the charging capacity of the electric automobile, P Work (work) Is the charging power.
The calculation formula of the charging quantity delta E of the electric automobile is as follows:
ΔE＝{SOC 2 -(SOC 1 +ΔSOC)}×C
SOC 2 battery charge and SOC for electric vehicle 1 The delta SOC is the battery charge quantity of the electric vehicle at the current position, delta SOC is the battery charge quantity loss of the electric vehicle from the current position to the initial optional charging station position, and C is the rated capacity of the electric vehicle battery.
P 1 : asset data is called and periodically updated according to prices formulated by national issuing and modifying committee;
P 2 : according to the pricing catalogue printed by the Beijing city issuing and modifying agent 2018, the pricing right of the charging service charge of the electric automobile is fully developed in 2018 and 4 months, and other provinces in the cultivation stage of the charging market are gradually released;
P 3 : retrieving asset data;
finally, the charging schemes with the top three scores are marked with different labels (such as optimal price, shortest time and the like) or key information (such as charging cost, idle rate and the like) is highlighted, and the key information is pushed to a user through the front end.
Through e user order data analysis display that charges earlier stage operation in-process obtained, the user can receive the state information of 5 charging stations through App when charging position decision is carried out, includes: charging station open time, idle rate, charging power, cost, and distance of the user's current location to the charging station. In the decision process, a user can estimate running time, queuing time, payment time and charging time to judge whether the opening time of a charging station is proper or not, estimate possible queuing waiting time through idle rate, estimate running time on the way through the distance from the current position to the charging station and intersection information on a map, estimate charging time through charging power of the charging station, and finally estimate charging cost through charging price (comprising electricity price and charging service charge) and parking cost.
A schematic diagram of a product prototype of the intelligent charging service recommendation system based on user images is shown in FIG. 5.
Example 2:
an infrastructure schematic diagram of an intelligent charging service recommendation system based on user images is shown in fig. 2, and includes:
an initial selectable charging station summoning module and a pushable charging station determining module;
the initial optional charging station summoning module is used for obtaining an initial optional charging station according to the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources;
the pushable charging station determining module is used for determining pushable charging stations from the initial selectable charging stations according to the pre-obtained user portrait tag weights;
the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data.
An intelligent charging service recommendation system detailed structure diagram based on user portraits is shown in fig. 3, and further comprises a user portrait tag weight training module, and the user portrait tag weight training module comprises:
the device comprises a label matching unit, a data processing unit and a data training unit;
the label base matching unit is used for carrying out label matching on the characteristic data and the user behavior data in the user order data according to the set user portrait label so as to generate a training log;
the data processing unit is used for processing the training log through a characteristic hash and low-frequency filtering method or an equal-frequency discretizing method to obtain training data;
the data training unit is used for bringing the training data into a multi-task deep neural network model for training to obtain user portrait tag weight;
the user portrait tag includes: open time, idle rate, distance, charge power, cost, and environment;
the characteristic data includes: user characteristic data, asset characteristic data, and environmental characteristic data.
The data processing unit comprises a characteristic hash subunit and a low-frequency filtering subunit;
the characteristic hash subunit is used for converting the training log into a real number matrix group through characteristic hash;
and the low-frequency filtering subunit is used for carrying out low-frequency filtering processing on the discrete features of the real number matrix group, removing the features smaller than the occurrence frequency threshold value and forming the training data.
Wherein, the initial optional charging station summoning module comprises: a safe driving range calculation unit, a safe driving range calculation unit and an initial selectable charging station determination unit;
the safe driving mileage calculation unit is used for calculating the safe driving mileage of the electric automobile in the current state;
the safe driving range calculation unit is used for determining the safe driving range of the electric automobile by taking the position of the electric automobile as a circle center and the safe driving mileage as a radius;
an initial optional charging station determining unit, configured to set charging stations in which all states are available in a safe driving range as initial optional charging stations;
the remaining driving range in the current state of the electric vehicle is determined by the safe driving range.
Wherein the pushable charging station determination module comprises: the mobile terminal comprises a user charging position decision influence factor value calculation unit, a pushable charging station determination unit and a pushing unit;
the user charging position decision influence factor value calculation unit is used for calculating user charging position decision influence factor values of all power stations in the initial selectable charging station;
the push-able charging station determining unit is used for determining push-able charging stations according to the pre-obtained user portrait tag weights and the user charging position decision influence factor values of the power stations;
the pushing unit is used for pushing the corresponding label on the pushing charging station to a client;
the user charging position decision influencing factor value of the power station comprises: time-impact factor value, cost-impact factor value, and environmental-impact factor value.
It should be noted that the above embodiments are only for illustrating the technical solution of the present application and not for limiting the scope of protection thereof, and although the present application has been described in detail with reference to the above embodiments, it should be understood by those skilled in the art that various changes, modifications or equivalents may be made to the specific embodiments of the application after reading the present application, and these changes, modifications or equivalents are within the scope of protection of the claims appended hereto.
Claims (16)
1. An intelligent charging service recommending method based on user images is characterized by comprising the following steps:
obtaining an initial charging station based on the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources;
determining a push charging station from the initial charging stations based on the pre-obtained user portrait tag weights;
the user portrait tag weight is based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data;
the determining a push charging station from the initial charging stations based on the pre-obtained user portrait tag weights comprises:
calculating a user charging position decision-making influence factor value of each power station in the initial charging station;
determining a pushing charging station based on a pre-obtained user portrait tag weight and a user charging position decision influence factor value of each power station;
the pushing charging station is marked with a corresponding label pushing client;
the user charging position decision influence factor value of the power station comprises: a time-impact factor value, a cost-impact factor value, and an environmental-impact factor value;
the user portrait tag weights include a time tag weight, a fee tag weight, and an environment tag weight; the time tag weight comprises an open time tag weight, an idle rate tag weight, a distance tag weight and a charging power tag weight.
2. The method of claim 1, wherein the user portrait tag weights include:
performing label matching on the characteristic data in the user order data and the user behavior data based on the set user portrait label to generate a training log;
processing the training log by a characteristic hash and low-frequency filtering method or an equal-frequency discretizing method to obtain training data;
the training data are put into a multi-task deep neural network model for training, and user portrait tag weights are obtained;
the user portrait tag includes: open time, idle rate, distance, charge power, cost, and environment;
the feature data includes: user characteristic data, asset characteristic data, and environmental characteristic data.
3. The method of claim 2, wherein the processing the training log by the feature hash and low frequency filtering method comprises:
converting the training log into a real number matrix group through characteristic hash;
and performing low-frequency filtering processing on the discrete features of the real number matrix group, and removing features smaller than an occurrence frequency threshold value to form the training data.
4. The method of claim 2, wherein the processing the training log by the equal frequency discretization method comprises:
and dividing the data in the training log according to the fixed frequency to obtain a plurality of groups of training data with equal sample size.
5. The method of claim 1, wherein the deriving an initial charging station based on an electric vehicle state, current coordinates of the electric vehicle, and charging station resource conditions comprises:
calculating the safe driving mileage of the electric automobile in the current state;
determining the safe driving range of the electric automobile by taking the current coordinate position of the electric automobile as a circle center and the safe driving mileage as a radius;
setting all charging stations with available states in the safe driving range of the electric automobile as initial charging stations;
the safety driving mileage of the electric automobile in the current state is determined by the remaining driving mileage.
6. The method of claim 5, wherein the remaining range calculation formula for the current state of the electric vehicle is as follows:
wherein S is rest For remaining mileage, SOC 1 And (3) the power battery charge quantity at the current coordinate position, SOH is the health state of the power battery, C is the rated capacity of the power battery, V is the current voltage of the power battery, and InitEC is the energy consumed by the unit driving mileage.
7. The method of claim 1, wherein the calculation formula for determining the push charging station is as follows:
Rank＝α·ΔT+β·P+γ·X
wherein Rank is an initial power station score, alpha is a time tag weight, beta is a cost tag weight, gamma is an environmental tag weight, deltaT is a time influence factor value, P is a cost influence factor value, and X is an environmental influence factor value.
8. The method of claim 7, wherein the time-dependent factor value is calculated as:
△T＝△T 2 +△T 3 +△T 4 +△T 5
wherein DeltaT is the time influence factor value DeltaT 2 For the travel time of the electric vehicle from the current position to the initial charging station position, deltaT 3 For queuing time after arrival at initial charging station, deltaT 4 Paying time for initial charging station, deltaT 5 The electric vehicle is charged for a time at an initial charging station.
9. The method of claim 8, wherein the electric vehicle is charged at an initial charging station for a time Δt 5 The calculation formula is as follows:
wherein delta E is the charging quantity of the electric automobile, and P Work (work) Is the charging power.
10. The method of claim 9, wherein the electric vehicle charge Δe is calculated as:
ΔE＝{SOC 2 -(SOC 1 +ΔSOC)}×C
wherein SOC is 2 Battery charge and SOC for electric vehicle 1 The delta SOC is the battery charge quantity of the electric vehicle at the current position, delta SOC is the battery charge quantity loss of the electric vehicle from the current position to the initial charging station position, and C is the rated capacity of the electric vehicle battery.
11. The method of claim 7, wherein the cost impact factor value P is calculated as:
P＝[P 1 +P 2 ]×△E+P 3 ×(△T 3 +△T 4 +△T 5 )；
wherein P is 1 Charging unit price, P for initial charging station 2 Charging service charge for initial charging station, delta E is charging quantity of electric automobile and P 3 For initial charging station parking fee, deltaT 3 For queuing time after arrival at initial charging station, deltaT 4 Paying time for initial charging station, deltaT 5 The electric vehicle is charged for a time at an initial charging station.
12. The method of claim 7, wherein the time tag weight α is calculated as:
α＝α 1 +α 2 +α 3 +α 4
wherein alpha is 1 Is the weight of the open time tag, alpha 2 For idle rate tag weight, alpha 3 Is the distance label weight alpha 4 The charge power is weighted.
13. An intelligent charging service recommendation system based on user images, for implementing the intelligent charging service recommendation method based on user images as claimed in claim 1, comprising:
an initial selectable charging station summoning module and a pushable charging station determining module;
the initial selectable charging station recall module is used for obtaining an initial selectable charging station according to the state of the electric vehicle, the current coordinates of the electric vehicle and the condition of charging station resources;
the pushed charging station determining module is used for determining the pushed charging station from the initial selectable charging stations according to the pre-obtained user portrait tag weight;
the user portrait tag weights are based on a multitasking deep neural network to train and determine feature data and user behavior data in user order data.
14. The system of claim 13, further comprising a user portrait tag weight training module, the user portrait tag weight training module comprising:
the device comprises a label matching unit, a data processing unit and a data training unit;
the label base matching unit is used for carrying out label matching on the characteristic data in the user order data and the user behavior data according to the set user portrait label so as to generate a training log;
the data processing unit is used for processing the training log through a characteristic hash and low-frequency filtering method or an equal-frequency discretizing method to obtain training data;
the data training unit is used for bringing the training data into a multi-task deep neural network model for training to obtain user portrait tag weights;
the user portrait tag includes: open time, idle rate, distance, charge power, cost, and environment;
the feature data includes: user characteristic data, asset characteristic data, and environmental characteristic data.
15. The system of claim 13, wherein the initial selectable charging station recall module comprises:
a safe driving range calculation unit, a safe driving range calculation unit and an initial selectable charging station determination unit;
the safe driving mileage calculation unit is used for calculating the safe driving mileage of the electric automobile in the current state;
the safe driving range calculation unit is used for determining the safe driving range of the electric automobile by taking the position of the electric automobile as a circle center and the safe driving mileage as a radius;
the initial selectable charging station determining unit is used for setting all charging stations with available states in the safe driving range as initial selectable charging stations;
the remaining driving mileage of the electric vehicle in the current state is determined by the safety driving mileage.
16. The system of claim 13, wherein the pushable charging station determination module comprises:
the mobile terminal comprises a user charging position decision influence factor value calculation unit, a pushable charging station determination unit and a pushing unit;
the user charging position decision influence factor value calculation unit is used for calculating user charging position decision influence factor values of all power stations in the initial selectable charging station;
the determination unit of the push-able charging station is used for determining the push-able charging station according to the pre-obtained user portrait tag weight and the user charging position decision influence factor value of each power station;
the pushing unit is used for pushing the corresponding label to the client on the pushing charging station;
the user charging position decision influence factor value of the power station comprises: time-impact factor value, cost-impact factor value, and environmental-impact factor value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911203017.6A CN111159533B (en) | 2019-11-29 | 2019-11-29 | Intelligent charging service recommendation method and system based on user image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911203017.6A CN111159533B (en) | 2019-11-29 | 2019-11-29 | Intelligent charging service recommendation method and system based on user image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111159533A CN111159533A (en) | 2020-05-15 |
| CN111159533B true CN111159533B (en) | 2023-10-20 |
Family
ID=70556305
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911203017.6A Active CN111159533B (en) | 2019-11-29 | 2019-11-29 | Intelligent charging service recommendation method and system based on user image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111159533B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111882096B (en) * | 2020-08-03 | 2022-11-29 | 博泰车联网科技(上海)股份有限公司 | Method for information processing, electronic device, and storage medium |
| CN112115350A (en) * | 2020-08-27 | 2020-12-22 | 上海电享信息科技有限公司 | Smart city charging intelligent recommendation method and system |
| CN112132608B (en) * | 2020-09-15 | 2024-03-19 | 北京慧辰资道资讯股份有限公司 | Charging service intelligent recommendation method and device based on big data of electric power Internet of things |
| CN112297934B (en) * | 2020-11-16 | 2022-08-12 | 国网北京市电力公司 | Charging station processing method and device, storage medium and processor |
| CN112700117B (en) * | 2020-12-29 | 2024-04-02 | 吉林大学 | Data-based electric vehicle charging infrastructure service condition estimation method |
| WO2023087108A1 (en) * | 2021-11-17 | 2023-05-25 | Datametrex Electric Vehicle Solutions Inc. | Systems, methods, and devices for determining optimal electric car charging stations |
| CN117172409B (en) * | 2023-09-05 | 2024-05-03 | 河海大学 | Intelligent charging method of electric automobile based on photovoltaic energy |
| CN116882715B (en) * | 2023-09-07 | 2023-11-28 | 杭州格创新能源有限公司 | Pile-car linkage orderly safe power utilization method and system based on cloud server |
| CN117689311B (en) * | 2023-12-04 | 2024-05-17 | 朴道征信有限公司 | Article inventory adjustment method, apparatus, electronic device, and computer-readable medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014038048A (en) * | 2012-08-17 | 2014-02-27 | Toshiba Corp | Charge management system |
| KR101439265B1 (en) * | 2014-02-10 | 2014-09-11 | 제주대학교 산학협력단 | Charging system and the method for electric vehicle |
| CN108304984A (en) * | 2017-10-27 | 2018-07-20 | 蔚来汽车有限公司 | Behavior quantization method and system |
| WO2018196745A1 (en) * | 2017-04-24 | 2018-11-01 | 知谷（上海）网络科技有限公司 | Charging control method and charging device |
| CN108829778A (en) * | 2018-05-30 | 2018-11-16 | 蔚来汽车有限公司 | Charging pile intelligent recommendation methods, devices and systems |
| CN108981736A (en) * | 2018-05-28 | 2018-12-11 | 国网江苏省电力有限公司电力科学研究院 | A kind of regular electric car charge path optimization method of being gone on a journey based on user |
| CN109501630A (en) * | 2018-12-04 | 2019-03-22 | 国网电动汽车服务有限公司 | A kind of electric car charging scheme real-time recommendation method and system |
| CN109559179A (en) * | 2017-09-26 | 2019-04-02 | 阿里巴巴集团控股有限公司 | Gas station's relevant information processing method and processing device |
| CN109858659A (en) * | 2017-11-30 | 2019-06-07 | 北京嘀嘀无限科技发展有限公司 | Charging pile website recommendation method, apparatus, server and readable storage medium storing program for executing |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110191186A1 (en) * | 2010-02-03 | 2011-08-04 | Levy Paul S | Method of autonomously recommending charging event of electric vehicle based on multiple factors and displaying carbon footprint data and personalized advertising via the vehicle display |
| US20150039391A1 (en) * | 2011-08-16 | 2015-02-05 | Better Place GmbH | Estimation and management of loads in electric vehicle networks |
| US9582748B2 (en) * | 2013-03-04 | 2017-02-28 | Hello Inc. | Base charging station for monitoring device |
| KR101676689B1 (en) * | 2016-03-23 | 2016-11-17 | 주식회사 아이온커뮤니케이션즈 | System and method for recommending charging station of electric vehicle |
-
2019
- 2019-11-29 CN CN201911203017.6A patent/CN111159533B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014038048A (en) * | 2012-08-17 | 2014-02-27 | Toshiba Corp | Charge management system |
| KR101439265B1 (en) * | 2014-02-10 | 2014-09-11 | 제주대학교 산학협력단 | Charging system and the method for electric vehicle |
| WO2018196745A1 (en) * | 2017-04-24 | 2018-11-01 | 知谷（上海）网络科技有限公司 | Charging control method and charging device |
| CN109559179A (en) * | 2017-09-26 | 2019-04-02 | 阿里巴巴集团控股有限公司 | Gas station's relevant information processing method and processing device |
| CN108304984A (en) * | 2017-10-27 | 2018-07-20 | 蔚来汽车有限公司 | Behavior quantization method and system |
| CN109858659A (en) * | 2017-11-30 | 2019-06-07 | 北京嘀嘀无限科技发展有限公司 | Charging pile website recommendation method, apparatus, server and readable storage medium storing program for executing |
| CN108981736A (en) * | 2018-05-28 | 2018-12-11 | 国网江苏省电力有限公司电力科学研究院 | A kind of regular electric car charge path optimization method of being gone on a journey based on user |
| CN108829778A (en) * | 2018-05-30 | 2018-11-16 | 蔚来汽车有限公司 | Charging pile intelligent recommendation methods, devices and systems |
| CN109501630A (en) * | 2018-12-04 | 2019-03-22 | 国网电动汽车服务有限公司 | A kind of electric car charging scheme real-time recommendation method and system |
Non-Patent Citations (2)
| Title |
|---|
| An Elastic Charging Service Fee-Based Load Guiding Strategy for Fast Charging Stations;Shu Su;Energies;全文 * |
| 智慧物联电动自行车充电系统;王欣;中国优秀硕士学位论文全文数据库;C034-320 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111159533A (en) | 2020-05-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111159533B (en) | Intelligent charging service recommendation method and system based on user image | |
| Ullah et al. | Electric vehicle energy consumption prediction using stacked generalization: An ensemble learning approach | |
| JP7381742B2 (en) | Battery service provision system and method | |
| Ullah et al. | Modeling of machine learning with SHAP approach for electric vehicle charging station choice behavior prediction | |
| CN105160428B (en) | The planing method of electric automobile on highway quick charge station | |
| CN110888908A (en) | Charging station/pile recommendation system and method capable of achieving deep learning continuously | |
| Yan et al. | Developmental trajectories of new energy vehicle research in economic management: Main path analysis | |
| CN106295860A (en) | A kind of electric automobile scale charge requirement Forecasting Methodology based on Monte Carlo Analogue Method | |
| CN108830414B (en) | Load prediction method for commercial charging area of electric automobile | |
| Zhang et al. | An improved adaptive equivalent consumption minimization strategy for parallel plug-in hybrid electric vehicle | |
| CN114039370A (en) | Resource optimization method for electric automobile and intelligent charging and discharging station based on V2G mode | |
| CN114282821A (en) | Scheduling method, system and equipment for sharing electric automobile | |
| Plagowski et al. | Impact of electric vehicle charging–An agent‐based approach | |
| Tian et al. | Method for predicting the remaining mileage of electric vehicles based on dimension expansion and model fusion | |
| Shi | Cobalt demand for automotive electrification in China: scenario analysis based on the bass model | |
| CN116008815A (en) | Method, device, equipment, storage medium and vehicle for detecting short circuit in battery cell | |
| Mishra et al. | Exploratory data analysis for electric vehicle driving range prediction: insights and evaluation | |
| Gu et al. | Big Data Mining Analysis of Key Indicators of Online New Energy Vehicles | |
| Salam et al. | Forecasting the Behavior of EV Charging Using Machine Learning | |
| CN112766665A (en) | Risk level-based electric charge recycling risk prevention and control method | |
| Shekari et al. | Recognition of electric vehicles charging patterns with machine learning techniques | |
| CN115204512B (en) | Vehicle charging behavior prediction model construction method based on data driving under multiple scenes | |
| Zhuo et al. | Digital-Twin-Driven Driving Range Prediction of Electric Vehicles | |
| Jin et al. | Electric vehicle charging demand forecast based on residents’ travel data | |
| CN107730108A (en) | A kind of highway fast charge network intelligence cloud service system and device |
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 |