CN110415462A - Atm device adds paper money optimization method and device - Google Patents
Atm device adds paper money optimization method and device Download PDFInfo
- Publication number
- CN110415462A CN110415462A CN201910700928.3A CN201910700928A CN110415462A CN 110415462 A CN110415462 A CN 110415462A CN 201910700928 A CN201910700928 A CN 201910700928A CN 110415462 A CN110415462 A CN 110415462A
- Authority
- CN
- China
- Prior art keywords
- paper money
- atm device
- data
- clear
- cycle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07F—COIN-FREED OR LIKE APPARATUS
- G07F19/00—Complete banking systems; Coded card-freed arrangements adapted for dispensing or receiving monies or the like and posting such transactions to existing accounts, e.g. automatic teller machines
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07F—COIN-FREED OR LIKE APPARATUS
- G07F19/00—Complete banking systems; Coded card-freed arrangements adapted for dispensing or receiving monies or the like and posting such transactions to existing accounts, e.g. automatic teller machines
- G07F19/20—Automatic teller machines [ATMs]
- G07F19/209—Monitoring, auditing or diagnose of functioning of ATMs
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Finance (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The present invention provides a kind of atm devices to add paper money optimization method and device, and atm device adds paper money optimization method to include: transaction data, geographical location, the data of weather forecast in the geographical location, paper money tankage, closing balance and the remaining clear number of days in the current clear cycle for obtain the atm device；According to data of weather forecast, paper money tankage, closing balance, the remaining clear number of days in the public holiday in the target clear cycle of the atm device, the transaction data in the current clear cycle, the geographical location, the target clear cycle, based on preset simulated environment model and add paper money planning optimization model, generates the atm device and paper money is added to operate.Atm device provided by the invention adds paper money optimization method and device that can quickly provide consider global cost plus paper money plan, reduces the operation expenditure of bank, provides more intelligent, more accurate, more efficient management means for bank's material object cash dosage management.
Description
Technical field
The present invention relates to artificial intelligence field technical fields, and in particular to the big data analysis field of financial industry, especially
It is to be related to a kind of atm device to add paper money plan optimization method and device.
Background technique
ATM self-help teller machine, which is business bank, provides the equipment of self-service access cash for client.Bank related service personnel
It need to estimate that every atm device stores cash amount according to portfolio and its related governing requirements, formulate cash to be administered equipment
Add paper money plan, carries out standby paper money and cash transfers.In the prior art, ATM cash management relies primarily on the artificial experience of business expert
The paper money amount of money, formulation cash is added to add paper money plan to estimate.What is considered by the formulation of cash plus paper money plan is global Optimum cost
And cash adds paper money plan to influence each other, cash adds paper money plan not only to need to meet administered ATM cash demand to want with clear management
It asks, also to balance the same day will add the future in paper money cost and next clear cycle to add paper money cost immediately.This causes cash to add paper money meter
The assessment complexity for the fine or not degree drawn is very big, and it is more excellent that artificial experience can not add paper money to pick out in the works from a large amount of feasible cashes
Cash add paper money plan.
Summary of the invention
For the problems of the prior art, atm device provided by the invention adds paper money optimization method and device that can quickly mention
Add paper money plan for the global cost of consideration, reduces the operation expenditure of bank, provide more intelligence for bank's material object cash dosage management
Energy, more accurate, more efficient management means.
In order to solve the above technical problems, the present invention the following technical schemes are provided:
In a first aspect, the present invention, which provides a kind of atm device, adds paper money optimization method, comprising:
Obtain the weather of transaction data in the current clear cycle of the atm device, geographical location, the geographical location
Forecast data, paper money tankage, closing balance and remaining clear number of days；
According to the public holiday in the target clear cycle of the atm device, the number of deals in the current clear cycle
According to data of weather forecast, paper money tankage, closing balance, the remaining clear day in, the geographical location, the target clear cycle
Number based on preset simulated environment model and adds paper money planning optimization model, generates the atm device and paper money is added to operate.
Preferably, the transaction data includes pipelined data, pay day and the repayment date of the atm device.
Preferably, atm device adds paper money optimization method further include:
Obtain transaction data in the history clear cycle of the atm device, geographical location, the day in history clear cycle
Destiny evidence, paper money tankage, closing balance, remaining clear number of days, clear cycle and the atm device target clear week obtained in advance
Prediction cash dosage in phase；
To transaction data, geographical location, the weather in history clear cycle in the history clear cycle of the atm device
Data, paper money tankage, closing balance, remaining clear number of days, clear cycle and the atm device target clear cycle obtained in advance
Interior prediction cash dosage is pre-processed；
It is clear according to the transaction data in the history clear cycle of the atm device after pretreatment, geographical location, history
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle and the ATM obtained in advance in the machine period are set
Prediction cash dosage in standby target clear cycle constructs the simulated environment model using supervised learning algorithm, wherein described
Simulated environment model is for training data needed for generating described plus paper money planning optimization model.
Preferably, the transaction data in the history clear cycle according to the atm device after pretreatment, geography
Position, the weather data in history clear cycle, paper money tankage, closing balance, remaining clear number of days, clear cycle and in advance
Prediction cash dosage in the atm device target clear cycle of acquisition constructs the simulated environment mould using supervised learning algorithm
Type, comprising:
By changing the atm device closing balance and remaining clear number of days, the supervised learning algorithm and described is utilized
Simulated environment model generates the simulated environment result under multiple simulated environments.
Preferably, the pretreatment includes: data scrubbing, data integration, data regularization and data transformation.
Preferably, atm device adds paper money optimization method further include:
The training data under different simulated environments is generated according to multiple simulated environment models；
Described plus paper money planning optimization model is generated using nitrification enhancement according to the training data under different simulated environments,
Wherein the nitrification enhancement includes: DQN algorithm, DDPG algorithm, IMPALA algorithm, A2C algorithm and GA3C algorithm.
Preferably, atm device adds paper money operation to include: that the atm device adds the paper money date and adds the paper money amount of money.
Second aspect, the present invention provide a kind of atm device and paper money are added to optimize device, which includes:
Data acquisition first unit, the transaction data in current clear cycle, geographical position for obtaining the atm device
It sets, the data of weather forecast in the geographical location, paper money tankage, closing balance and remaining clear number of days；
Add paper money to operate generation unit, in the target clear cycle according to the atm device public holiday, described work as
Transaction data, the geographical location in preceding clear cycle, the data of weather forecast in the target clear cycle, paper money case hold
Amount, closing balance, remaining clear number of days based on preset simulated environment model and add paper money planning optimization model, generate the ATM
Equipment adds paper money to operate.
Preferably, the transaction data includes pipelined data, pay day and the repayment date of the atm device.
Preferably, atm device adds paper money to optimize device further include:
Data acquisition second unit, the transaction data in history clear cycle, geographical position for obtaining the atm device
It sets, the weather data in history clear cycle, paper money tankage, closing balance, remaining clear number of days, clear cycle and obtain in advance
The prediction cash dosage in atm device target clear cycle taken；
Pretreatment unit, for the transaction data in the history clear cycle to the atm device, geographical location, history
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle in clear cycle and the ATM obtained in advance
Prediction cash dosage in device target clear cycle is pre-processed；
Environment geometric modeling unit, for the friendship in the history clear cycle according to the atm device after pretreatment
Easy data, geographical location, the weather data in history clear cycle, paper money tankage, closing balance, remaining clear number of days, clear
Prediction cash dosage in period and the atm device target clear cycle obtained in advance constructs institute using supervised learning algorithm
State simulated environment model, wherein the simulated environment model is for training number needed for generating described plus paper money planning optimization model
According to.
Preferably, environment geometric modeling unit is specifically used for by changing the atm device closing balance and remaining clear
Number of days generates the simulated environment result under multiple simulated environments using the supervised learning algorithm and the simulated environment model.
Preferably, the pretreatment unit is specifically used for data scrubbing, data integration, data regularization and data transformation.
Preferably, atm device adds paper money to optimize device further include:
Training data generation unit, for generating the training number under different simulated environments according to multiple simulated environment models
According to；
Add paper money model generation unit, for generating according to the training data under different simulated environments using nitrification enhancement
Described plus paper money planning optimization model, wherein the nitrification enhancement includes: DQN algorithm, DDPG algorithm, IMPALA algorithm, A2C
Algorithm and GA3C algorithm.
The third aspect, the present invention provides a kind of electronic equipment, including memory, processor and storage are on a memory and can
The computer program run on a processor realizes the step of atm device adds paper money optimization method when processor executes program.
Fourth aspect, the present invention provide a kind of computer readable storage medium, are stored thereon with computer program, the calculating
The step of atm device adds paper money optimization method is realized when machine program is executed by processor.
As can be seen from the above description, a kind of atm device of the present invention adds paper money optimization method and device, this method includes simulated environment
Foundation, the training of intensified learning model, the cash of intensified learning model add paper money plan generation, the evaluation of system at regular intervals feedback.
Specifically, it before progress ATM adds paper money to plan, needs using simulated environment model foundation simulated environment.Paper money planning optimization model is added to exist
It is constantly interacted with simulated environment when intensified learning model training, the update of model is carried out while generating a large amount of training datas
Training.Training is until by assessment, model training process terminates.
To sum up, atm device provided by the invention add paper money optimization method comprehensively consider plus the transportation cost of paper money, clear paper money cost,
Cash accounts for money cost, atm management requirement is designed and suitably adds paper money cost calculation mode, and intensified learning model characteristics is utilized to balance
Instant cost and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan,
Effectively promoted the management of cash in bank dosage precision and working efficiency, thus reduce bank's inventory cash total amount and O&M at
This, is greatly decreased business personnel's work load, promotes site cash service coverage ratio.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is the present invention
Some embodiments for those of ordinary skill in the art without creative efforts, can also basis
These attached drawings obtain other attached drawings.
Fig. 1 is that atm device adds paper money optimization method flow diagram one in the embodiment of the present invention；
Fig. 2 is transaction data composition schematic diagram in the embodiment of the present invention；
Fig. 3 is that atm device adds paper money optimization method flow diagram two in the embodiment of the present invention；
Fig. 4 is to pre-process composition schematic diagram in the embodiment of the present invention；
Fig. 5 is that atm device adds paper money optimization method flow diagram three in the embodiment of the present invention；
Fig. 6 is that atm device adds paper money to operate composition schematic diagram in the embodiment of the present invention；
Fig. 7 is the flow diagram that atm device adds paper money optimization method in specific application example of the invention；
Fig. 8 interacts schematic diagram with paper money planning optimization model is added for simulated environment model in specific application example of the invention；
Fig. 9 is the flow diagram that atm device adds step S5 in paper money optimization method in specific application example of the invention；
Figure 10 is the structural schematic diagram one that atm device adds paper money optimization device in specific application example of the invention；
Figure 11 is the structural schematic diagram two that atm device adds paper money optimization device in specific application example of the invention；
Figure 12 is the structural schematic diagram three that atm device adds paper money optimization device in specific application example of the invention；
Figure 13 is the structural schematic diagram of the electronic equipment in the embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, the technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
The embodiment of the present invention provides the specific embodiment that a kind of atm device adds paper money optimization method, referring to Fig. 1 this method10Specifically include following content:
Step 100: obtaining transaction data in the current clear cycle of the atm device, geographical location, the geographical position
Data of weather forecast, paper money tankage, closing balance and the remaining clear number of days set.
It is understood that clear is exactly the paper money case emptied in atm device.Mainly by with elder generation and backstage reconciliation, will
Paper money case is brought back clearing subbranch and is checked.Clear further includes taking flowing water paper, takes retain card etc..Ordinary circumstance is all to make clear to add paper money, is taken old
Paper money case renews paper money case, change of current water paper etc..
It should be pointed out that current clear cycle can be multiple periods such as one day, one week, one month following.Step 100
In transaction data include atm device pipelined data, pay day and repayment date.
Step 200: according in the target clear cycle of the atm device public holiday, in the current clear cycle
The transaction data, geographical location, the data of weather forecast in the target clear cycle, paper money tankage, closing balance, surplus
Remaining clear number of days based on preset simulated environment model and adds paper money planning optimization model, generates the atm device and paper money is added to operate.
It is understood that the simulated environment model is for training number needed for generating described plus paper money planning optimization model
According to simulated environment model using TSP approximate algorithm, (calculate by approximate algorithm Prim algorithm, such as Hopfield neural network algorithm, heredity
Method, simulated annealing, ant group algorithm, tabu search algorithm and greedy algorithm) it is calculated.
In the public holiday in the target clear cycle by the atm device, the number of deals in the current clear cycle
According to data of weather forecast, paper money tankage, closing balance, the remaining clear day in, the geographical location, the target clear cycle
After number is by preset simulated environment model treatment, the training data under different simulated environments is generated, and is input to and adds
In paper money planning optimization model, ultimately generates the atm device and paper money is added to operate.
Atm device provided by the invention adds that paper money optimization method comprehensively considers plus transportation cost, clear paper money cost, the cash of paper money accounts for
Money cost, atm management require design suitably plus paper money cost calculation mode, using intensified learning model characteristics balance immediately at
Sheet and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan, effectively mention
The precision and working efficiency for rising the management of cash in bank dosage, to reduce bank's inventory cash total amount and O&M cost, substantially
Business personnel's work load is reduced, site cash service coverage ratio is promoted.
In one embodiment, referring to fig. 2, the transaction data includes pipelined data, pay day and the refund of the atm device
Day.
It is understood that the transaction data includes pipelined data, pay day and the repayment date of the atm device.
In one embodiment, referring to Fig. 3, atm device adds paper money optimization method further include:
Step 300: transaction data, geographical location, the history clear obtained in the history clear cycle of the atm device is all
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle in phase and the atm device mesh obtained in advance
Mark the prediction cash dosage in clear cycle.
Step 400: to the transaction data in the history clear cycle of the atm device, geographical location, history clear cycle
Interior weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle and the atm device target obtained in advance
Prediction cash dosage in clear cycle is pre-processed.
Pretreatment in step 400 includes data scrubbing, data integration, data regularization and data transformation.
Step 500: according to the transaction data in the history clear cycle of the atm device after pretreatment, geographical position
It sets, the weather data in history clear cycle, paper money tankage, closing balance, remaining clear number of days, clear cycle and obtain in advance
The prediction cash dosage in atm device target clear cycle taken constructs the simulated environment model using supervised learning algorithm.
Wherein, the simulated environment model is for training data needed for generating described plus paper money planning optimization model.
It is understood that simulated environment model can for plus paper money planning optimization model the available instruction of a large amount of high quality is provided
Practice data.In addition, supervised learning algorithm refers to the parameter of the sample adjustment classifier using one group of known class, institute is reached
It is required that the process of performance, also referred to as supervised training or there is teacher learning.Specifically, supervised learning algorithm is the training number from label
According to come the machine learning task of inferring a function.Training data includes a set of training example.In supervised learning, each example
All it is made of an input object (usually vector) and a desired output valve (also referred to as supervisory signals).Supervised learning
Algorithm is to analyze the training data, and generate the function of a deduction, can be used for mapping out new example.One optimal
Scheme correctly determines the class label of those invisible examples by the algorithm is allowed.This requires learning algorithm is in one kind
The mode of " reasonable " is formed under from training data to invisible situation from one kind.
In one embodiment, step 500 can not also be realized by supervised learning algorithm, i.e., directly using according to pretreatment
The day destiny in transaction data, geographical location, history clear cycle in the history clear cycle of the atm device later
According to, paper money tankage, closing balance, remaining clear number of days, clear cycle and in advance in the atm device target clear cycle that obtains
Prediction cash dosage generate simulated environment model.Specifically, original state (the ATM closing balance of simulated environment model is modified
With remaining clear cycle), so that simulated environment model is recalculated corresponding time of historical data according to transaction data, to obtain
Model environment model is directed under different environment (different adds paper money plan), has what kind of result (cost consumption).Having
When body is implemented, artificial simulation and forecast by the way of increasing noise is combined using true sale data.Wherein the simulation of noise is basis
The mean bias of supervised learning forecast result of model and variance obtain in history.
It, can be with it is understood that reduce coupling The method avoids directly relying on to supervised learning model
Result of the supervised learning model after multiple optimization is set to be directly used in intensified learning model.It avoids modifying supervised learning model every time
The needs of intensified learning model are newly trained afterwards, (supervised learning model refers to ATM cash forecast dosage).
In one embodiment, step 500 specifically: by changing the atm device closing balance and remaining clear number of days, benefit
The simulated environment result under multiple simulated environments is generated with the supervised learning algorithm and the simulated environment model.
It is understood that simulated environment model is responsible for integral data building simulated environment.ATM Transaction Information, ATM clear
Period, ATM paper money tankage are objective ATM use environment data, are necessary to ensure that and do not change in simulated environment.Simulated environment mould
Block can simulate different ATM and add paper money planning environment by the way that different ATM closing balances, ATM residue clear number of days is arranged.
In one embodiment, referring to fig. 4, the pretreatment includes: that data scrubbing, data integration, data regularization and data become
It changes.
In one embodiment, referring to Fig. 5, atm device adds paper money optimization method further include:
Step 600: the training data under different simulated environments is generated according to multiple simulated environment models.
Step 700: described plus paper money plan is generated using nitrification enhancement according to the training data under different simulated environments
Optimized model.
It is understood that the nitrification enhancement in step 700 is a kind of trial and error learning algorithm.In nitrification enhancement
In the task of application, tutorial message is seldom, and is often just providing afterwards.Due to performing nothing after some is acted in task
Such issues that method directly judges whether it is suitably to act, and the method for supervised learning can not solve.And in nitrification enhancement
In, model can be interacted constantly with environment, obtain optimal strategy by way of trial and error.Wherein the intensified learning is calculated
Method includes: DQN algorithm, DDPG algorithm, IMPALA algorithm, A2C algorithm and GA3C algorithm.
In one embodiment, referring to Fig. 6, atm device adds paper money operation to include: that the atm device adds the paper money date and adds the paper money amount of money.
Specifically, add paper money operation to refer to need when to carry out that paper money and determination is added to add the paper money amount of money.
As can be seen from the above description, a kind of atm device of the present invention adds paper money optimization method, this method includes building for simulated environment
The vertical, training of intensified learning model, intensified learning model cash add paper money plan generation, the evaluation of system at regular intervals feedback.Specifically
Ground needs before progress ATM adds paper money to plan using simulated environment model foundation simulated environment.Paper money planning optimization model is added to strengthen
It is constantly interacted with simulated environment when learning model training, the update instruction of model is carried out while generating a large amount of training datas
Practice.Training is until by assessment, model training process terminates.
To sum up, atm device provided by the invention add paper money optimization method comprehensively consider plus the transportation cost of paper money, clear paper money cost,
Cash accounts for money cost, atm management requirement is designed and suitably adds paper money cost calculation mode, and intensified learning model characteristics is utilized to balance
Instant cost and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan,
Effectively promoted the management of cash in bank dosage precision and working efficiency, thus reduce bank's inventory cash total amount and O&M at
This, is greatly decreased business personnel's work load, promotes site cash service coverage ratio.
To further explain this programme, the present invention provides the specific application example that atm device adds paper money optimization method, the tool
Body application example specifically includes following content20, referring to Fig. 7.
S0: data are obtained.
Required data include ATM Transaction Information, ATM paper money tankage, ATM closing balance, ATM residue clear number of days, ATM clear
Machine period and ATM cash forecast model be 1 day following to ATM, 2 days, 3 days, 4 days, 5 days, 6 days, 7 days, 8 days totally eight clear weeks
Prediction cash demand amount of phase etc. influences the data that cash adds paper money plan.And 8 prediction cash demand amounts are carried out respectively
Data processing finds out the variance of all data, while calculating the mean error of all data and true value, in simulated environment
Noise.
S1: building simulated environment model.
Specifically, by pretreatment after atm device history clear cycle in transaction data, geographical location,
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle and acquisition in advance in history clear cycle
Atm device target clear cycle in prediction cash dosage, utilize supervised learning algorithm to construct the simulated environment model.Tool
Body, keeps objective data outside business, such as ATM transaction data, date are constant, pass through modification business internal data, such as ATM
Closing balance, ATM residue clear number of days etc. construct simulated environment.It is understood that simulated environment can fill for model training
It sets and the available training data of a large amount of high quality is provided.
It is understood that simulated environment model is responsible for the current shape for adding paper money planning optimization model offer simulated environment
State.During adding paper money planning optimization model training, needs simulated environment model to provide data and support.To the state letter of simulated environment
Breath carries out summarizing calculating, finally environmental state information needed for adding paper money planning optimization model is spread out of, environmental state information includes ATM
That clear cycle, cash add paper money plan plus paper money path length, ATM paper money tankage and same day ATM closing balance subtract 1 day, 2
It, 3 days, 4 days, 5 days, 6 days, 7 days, the obtained ATM of prediction cash demand amount following 8 days of 8 days totally eight clear cycles it is every
It prediction closing balance etc..Wherein, cash add paper money plan plus the calculating of paper money path length belong to TSP problem, when exact algorithm
Between complexity is high, be difficult to the short time in obtain accurately plus paper money path length.Therefore, simulated environment model uses approximate algorithm
Prim algorithm is calculated.Since the analog form of prediction cash demand amount is calculated really now by statistics ATM Transaction Information
Golden demand adds in conjunction with historical forecast variance (current balance combines 8 days in the future historical trading datas to be calculated) with error
Enter noise (obtaining in step S0) to be obtained, the producible data volume of the simulation of simulated environment model each round is than original available number
It is 8 days few according to length.
S2: obtaining plus paper money operating cost calculation formula.
Specifically, it is interacted with business personnel, the weight of procurement cost influence factor, determines that cost uses average weighted side
Formula calculates.It should be pointed out that after the interaction with virtual environment each time, require to calculate cost price when training pattern,
The cost impact fed back in the weight of cost price cost impact factor as set by business personnel and simulated environment model because
Element weighted average gained.Wherein, cost impact factor includes but is not limited to add paper money path length, adds paper money number, counting handling
Money is accounted for, cash, lacks paper money number, clear cycle.
S3: using nitrification enhancement generates and train add paper money planning optimization model.
Here nitrification enhancement principle is illustrated by taking chess game as an example, computer chess player is not aware that in next step when starting
Chess be to be it is wrong, less know which step chess is the key that whole chess game office gets the upper hand of, but is known that according to result of the match and walks out this
After walking chess (final step for referring to chess game), the result is that defeated or win, if walked so last the result is that triumph, algorithm
(computer chess player) with regard to learning and memory, if last defeated, algorithm is no longer moved as this disk chess after just learning.
Described plus paper money planning optimization model is generated using DDPG algorithm according to the training data under different simulated environments.DQN
It is a kind of model free (no environmental model), off-policy (strategy for generating the strategy of behavior and being assessed is different)
Nitrification enhancement.DDPG (Deep Deterministic Policy Gradient) algorithm is also model free,
Off-policy's or on-policy, and equally used deep neural network for approximation to function.But unlike DQN,
DQN can only solve the problems, such as the not high action spaces of discrete and dimension, and this point please recall the defeated of the neural network of DQN
Out.And DDPG can solve Continuous action space problem.In addition, DQN is value based method, i.e. only one value function
Network, and DDPG is actor-critic method, i.e., existing value function network (critic), and has tactful network (actor).It can
With understanding, intensified learning is the process to iterate, and iteration will solve two problems each time: giving a strategy
Value finding function, and according to value function come more new strategy.
Carry out approximate value functions using a neural network in DDPG, this value function network is also known as critic network, it defeated
Entering is action and observation [a, s], output be Q (s, a)；In addition carry out approximate strategic function using a neural network,
This policy network is also known as actor network, its input is observation ss, and output is action a.
Specifically, actor:a=π (s；θ) a=π (s；θ)
Target actor:a=π (s；θ -) a=π (s；θ-)
It should be noted that "-" in above formula be the actor network that upper bout updates parameter.Wherein " bout "
It refers to that algorithm updates the round of target network, the use of algorithm is that a value can be set, whenever actor and critic network changes
After certain number, their parameter value can be covered the parameter of a target network, at this moment target network and current
Actor with critic network is consistent.
Connection between two networks is as described below: environment first can provide an obs, intelligent body according to actor network (after
Face, which can be talked about, increases noise on this network foundation) make a policy action, and environment can provide a prize after receiving this action
Encourage Rew and new obs.This process is a step.It is gone to update critic network according to Rew at this time, then be built along critic
The direction of view is gone to update actor network.Subsequently enter next step.So circulation is gone down, good until having trained one
Actor network.
Target network is also used as DQN, in DDPG to guarantee the convergence of parameter.Assuming that critic network is Q
(s,a；ω)Q(s,a；ω), its corresponding target critic network is Q (s, a；ω-)Q(s,a；ω-).Actor network is
π(s；θ)π(s；θ), its corresponding target actor network is π (s；θ-)π(s；θ-).
Critic network is approximate for value function.
Targett=Rt+1+ γ Q (St+1, π (St+1；θ-)；ω -) targett=Rt+1+ γ Q (St+1, π (St+1；
θ-)；ω-)
Loss=1NN ∑ t=1 (targett-Q (St, at；ω)) 2Loss=1N ∑ t=1N (targett-Q (St, at；
ω))2
Then it is updated using gradient descent method.Note that actor and critic all employ target network to calculate
target.Actor network is used for parameterization method.Refer here to a very important concept in intensified learning: strategy ladder
Spend Policy Gradient.
In the training process, add paper money planning optimization model to interact with simulated environment model, generated during interaction big
Measure training data.Need to assess interactive action after interaction every time simultaneously, i.e. cash adds paper money operating cost.It constantly updates and adds paper money plan
The inner parameter of Optimized model.When every model training for carrying out certain round, simulation is strengthened in simulated environment plus paper money plan is excellent
Change the actual use situation of model.After the complete interaction of a wheel by simulated environment, assessment factor of the statistics when wheel interaction.If commenting
Estimate factor then terminates to train by preset evaluation condition, otherwise continues to train.Use experience playback may be selected in training to calculate,
While to guarantee training speed, the interaction with simulated environment model is reduced.
Referring to Fig. 8, during model training simulated environment model with plus paper money planning optimization model interactive process it is as follows:
1. plus paper money planning optimization model needs to obtain Observation environmental state information to simulated environment model.
2. plus environmental state information input simulated environment model is obtained Action cash and adds paper money meter by paper money planning optimization model
It draws.
Paper money plan is added to be updated environment 3. simulated environment model receives cash by environmental renewal.
4. after environmental renewal, the environmental feedback module of simulated environment model counts again calculates ambient condition.
5. this cash of the movement assessment of cost module estimation of paper money planning optimization model is added to add paper money while continuing to execute 1.
The Reward cost of plan.
S4: it generates atm device and adds paper money plan.
Add paper money planning optimization model to generate cash in true environment (truthful data) and adds paper money plan.Specifically, it obtains true
Real environment data are input to and add paper money planning optimization model, add paper money plan to generate cash.
S5: evaluation plus paper money plan.
Current plus paper money planning optimization forecast result of model is evaluated and fed back every prefixed time interval.According to feedback
Decide whether adjusting training.Specifically, include: referring to Fig. 9, step S5
Step S501 obtains all prediction data till now from last time evaluation feedback.
Step S502, statistical service personnel assessment factor of interest are assessed.
Step S503 obtains business personnel to the empirical evaluation of evaluation factor result, while business personnel being allowed to mention
For another subjective model evaluation table.
Step S504, the objective evaluation factor effect of statistics, business personnel subjective assessment table and empirical evaluation.It is periodically right
Currently plus paper money planning optimization forecast result of model is evaluated, and when there is any one poor, is evaluated according to evaluation height modification
Threshold value (more stringent), and carry out continuing to train using full dose data.It is understood that plus paper money planning optimization model with
Time change and the environment located constantly changes, therefore when evaluation result is preferable, still need to be trained using full dose data.
As can be seen from the above description, a kind of atm device of the present invention adds paper money optimization method, this method includes building for simulated environment
The vertical, training of intensified learning model, intensified learning model cash add paper money plan generation, the evaluation of system at regular intervals feedback.Specifically
Ground needs before progress ATM adds paper money to plan using simulated environment model foundation simulated environment.Paper money planning optimization model is added to strengthen
It is constantly interacted with simulated environment when learning model training, the update instruction of model is carried out while generating a large amount of training datas
Practice.Training is until by assessment, model training process terminates.
To sum up, atm device provided by the invention add paper money optimization method comprehensively consider plus the transportation cost of paper money, clear paper money cost,
Cash accounts for money cost, atm management requirement is designed and suitably adds paper money cost calculation mode, and intensified learning model characteristics is utilized to balance
Instant cost and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan,
Effectively promoted the management of cash in bank dosage precision and working efficiency, thus reduce bank's inventory cash total amount and O&M at
This, is greatly decreased business personnel's work load, promotes site cash service coverage ratio.
Based on the same inventive concept, the embodiment of the present application also provides atm devices plus paper money to optimize device, can be used to implement
Method described in above-described embodiment, such as the following examples.Due to atm device plus the paper money principle that solves the problems, such as of optimization device and
Atm device adds paper money optimization method similar, therefore atm device adds the implementation of paper money optimization device to may refer to atm device paper money is added to optimize
Method is implemented, and overlaps will not be repeated.Used below, predetermined function may be implemented in term " unit " or " module "
The combination of software and/or hardware.Although system described in following embodiment is preferably realized with software, hardware, or
The realization of the combination of person's software and hardware is also that may and be contemplated.
The embodiment of the present invention provide it is a kind of can be realized atm device add the atm device of paper money optimization method add paper money optimization dress
The specific embodiment set, referring to Figure 10, atm device adds paper money optimization device to specifically include following content:
Data acquisition first unit 10, the transaction data in current clear cycle, geography for obtaining the atm device
Position, the data of weather forecast in the geographical location, paper money tankage, closing balance and remaining clear number of days；
Paper money is added to operate generation unit 20, for public holiday, described in the target clear cycle according to the atm device
Transaction data, the geographical location, the data of weather forecast in the target clear cycle, paper money case in current clear cycle hold
Amount, closing balance, remaining clear number of days based on preset simulated environment model and add paper money planning optimization model, generate the ATM
Equipment adds paper money to operate.
Preferably, the transaction data includes pipelined data, pay day and the repayment date of the atm device.
Preferably, referring to Figure 11, atm device adds paper money to optimize device further include:
Data acquisition second unit 30, the transaction data in history clear cycle, geography for obtaining the atm device
Position, the weather data in history clear cycle, paper money tankage, closing balance, remaining clear number of days, clear cycle and in advance
Prediction cash dosage in the atm device target clear cycle of acquisition；
Pretreatment unit 40, in the history clear cycle to the atm device transaction data, geographical location, go through
It weather data, paper money tankage, closing balance in history clear cycle, remaining clear number of days, clear cycle and obtains in advance
Prediction cash dosage in atm device target clear cycle is pre-processed；
Environment geometric modeling unit 50, in the history clear cycle according to the atm device after pretreatment
It is transaction data, geographical location, the weather data in history clear cycle, paper money tankage, closing balance, remaining clear number of days, clear
Prediction cash dosage in machine period and the atm device target clear cycle obtained in advance, is constructed using supervised learning algorithm
The simulated environment model, wherein the simulated environment model is for training needed for generating described plus paper money planning optimization model
Data.
Preferably, environment geometric modeling unit is specifically used for by changing the atm device closing balance and remaining clear
Number of days generates the simulated environment result under multiple simulated environments using the supervised learning algorithm and the simulated environment model.
Preferably, the pretreatment unit is specifically used for data scrubbing, data integration, data regularization and data transformation.
Preferably, referring to Figure 12, atm device adds paper money to optimize device further include:
Training data generation unit 60, for generating the training number under different simulated environments according to multiple simulated environment models
According to；
Add paper money model generation unit 70, for raw using nitrification enhancement according to the training data under different simulated environments
At described plus paper money planning optimization model, wherein the nitrification enhancement includes: DQN algorithm, DDPG algorithm, IMPALA algorithm,
A2C algorithm and GA3C algorithm.
As can be seen from the above description, a kind of atm device of the present invention adds paper money to optimize device, this method includes building for simulated environment
The vertical, training of intensified learning model, intensified learning model cash add paper money plan generation, the evaluation of system at regular intervals feedback.Specifically
Ground needs before progress ATM adds paper money to plan using simulated environment model foundation simulated environment.Paper money planning optimization model is added to strengthen
It is constantly interacted with simulated environment when learning model training, the update instruction of model is carried out while generating a large amount of training datas
Practice.Training is until by assessment, model training process terminates.
To sum up, atm device provided by the invention add paper money optimization device comprehensively consider plus the transportation cost of paper money, clear paper money cost,
Cash accounts for money cost, atm management requirement is designed and suitably adds paper money cost calculation mode, and intensified learning model characteristics is utilized to balance
Instant cost and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan,
Effectively promoted the management of cash in bank dosage precision and working efficiency, thus reduce bank's inventory cash total amount and O&M at
This, is greatly decreased business personnel's work load, promotes site cash service coverage ratio.
Embodiments herein also provides the atm device that can be realized in above-described embodiment and adds all to be walked in paper money optimization method
The specific embodiment of rapid a kind of electronic equipment, referring to Figure 13, electronic equipment specifically includes following content:
Processor (processor) 1201, memory (memory) 1202, communication interface (Communications
Interface) 1203 and bus 1204；
Wherein, processor 1201, memory 1202, communication interface 1203 complete mutual communication by bus 1204；
Communication interface 1203 passes for realizing the information between the relevant devices such as server-side devices, recording equipment and ustomer premises access equipment
It is defeated.
Processor 1201 is used to call the computer program in memory 1202, and processor is realized when executing computer program
Atm device in above-described embodiment adds the Overall Steps in paper money optimization method, for example, processor is realized when executing computer program
Following step:
Step 100: obtaining transaction data in the current clear cycle of the atm device, geographical location, the geographical position
Data of weather forecast, paper money tankage, closing balance and the remaining clear number of days set；
Step 200: according in the target clear cycle of the atm device public holiday, in the current clear cycle
The transaction data, geographical location, the data of weather forecast in the target clear cycle, paper money tankage, closing balance, surplus
Remaining clear number of days based on preset simulated environment model and adds paper money planning optimization model, generates the atm device and paper money is added to operate.
As can be seen from the above description, the electronic equipment in the embodiment of the present application, this method includes the foundation of simulated environment, strengthens
The training of learning model, the cash of intensified learning model add paper money plan generation, the evaluation of system at regular intervals feedback.Specifically, it carries out
Before ATM adds paper money to plan, need using simulated environment model foundation simulated environment.Add paper money planning optimization model in intensified learning model
It is constantly interacted with simulated environment when training, the update training of model is carried out while generating a large amount of training datas.Training is straight
To by assessment, model training process terminates.
To sum up, atm device provided by the invention add paper money optimization method comprehensively consider plus the transportation cost of paper money, clear paper money cost,
Cash accounts for money cost, atm management requirement is designed and suitably adds paper money cost calculation mode, and intensified learning model characteristics is utilized to balance
Instant cost and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan,
Effectively promoted the management of cash in bank dosage precision and working efficiency, thus reduce bank's inventory cash total amount and O&M at
This, is greatly decreased business personnel's work load, promotes site cash service coverage ratio.
Embodiments herein also provides the atm device that can be realized in above-described embodiment and adds all to be walked in paper money optimization method
A kind of rapid computer readable storage medium is stored with computer program on computer readable storage medium, the computer program
Realize that the atm device in above-described embodiment adds the Overall Steps of paper money optimization method when being executed by processor, for example, processor executes
Following step is realized when computer program:
Step 100: obtaining transaction data in the current clear cycle of the atm device, geographical location, the geographical position
Data of weather forecast, paper money tankage, closing balance and the remaining clear number of days set；
Step 200: according in the target clear cycle of the atm device public holiday, in the current clear cycle
The transaction data, geographical location, the data of weather forecast in the target clear cycle, paper money tankage, closing balance, surplus
Remaining clear number of days based on preset simulated environment model and adds paper money planning optimization model, generates the atm device and paper money is added to operate.
As can be seen from the above description, the computer readable storage medium in the embodiment of the present application, this method includes simulated environment
Foundation, the training of intensified learning model, the cash of intensified learning model add paper money plan generation, the evaluation of system at regular intervals feedback.
Specifically, it before progress ATM adds paper money to plan, needs using simulated environment model foundation simulated environment.Paper money planning optimization model is added to exist
It is constantly interacted with simulated environment when intensified learning model training, the update of model is carried out while generating a large amount of training datas
Training.Training is until by assessment, model training process terminates.
To sum up, atm device provided by the invention add paper money optimization method comprehensively consider plus the transportation cost of paper money, clear paper money cost,
Cash accounts for money cost, atm management requirement is designed and suitably adds paper money cost calculation mode, and intensified learning model characteristics is utilized to balance
Instant cost and future cost, are automatically performed the formulation work that ATM cash adds paper money plan, provide preferably cash and add paper money plan,
Effectively promoted the management of cash in bank dosage precision and working efficiency, thus reduce bank's inventory cash total amount and O&M at
This, is greatly decreased business personnel's work load, promotes site cash service coverage ratio.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for hardware+
For program class embodiment, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to side
The part of method embodiment illustrates.
It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims
It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment
It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable
Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can
With or may be advantageous.
Although this application provides the method operating procedure of such as embodiment or flow chart, based on routine or without creativeness
Labour may include more or less operating procedure.The step of enumerating in embodiment sequence is only that numerous steps execute
One of sequence mode, does not represent and unique executes sequence.It, can be by when device in practice or client production execute
It is executed according to embodiment or method shown in the drawings sequence or parallel executes (such as parallel processor or multiple threads
Environment).
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Specific embodiment is applied in the present invention, and principle and implementation of the present invention are described, above embodiments
Explanation be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art,
According to the thought of the present invention, there will be changes in the specific implementation manner and application range, in conclusion in this specification
Appearance should not be construed as limiting the invention.
Claims (15)
1. a kind of atm device adds paper money optimization method characterized by comprising
Obtain the weather forecast of the transaction data, geographical location, the geographical location in the current clear cycle of the atm device
Data, paper money tankage, closing balance and remaining clear number of days；
According to the public holiday in the target clear cycle of the atm device, the transaction data in the current clear cycle, institute
State geographical location, the data of weather forecast in the target clear cycle, paper money tankage, closing balance, remaining clear number of days, base
In preset simulated environment model and add paper money planning optimization model, generates the atm device and paper money is added to operate.
2. atm device according to claim 1 adds paper money optimization method, which is characterized in that the transaction data includes described
Pipelined data, pay day and the repayment date of atm device.
3. atm device according to claim 2 adds paper money optimization method, which is characterized in that further include:
Obtain transaction data in the history clear cycle of the atm device, geographical location, the day destiny in history clear cycle
According to, paper money tankage, closing balance, remaining clear number of days, clear cycle and in advance in the atm device target clear cycle that obtains
Prediction cash dosage；
To the transaction data in the history clear cycle of the atm device, geographical location, the day destiny in history clear cycle
According to, paper money tankage, closing balance, remaining clear number of days, clear cycle and in advance in the atm device target clear cycle that obtains
Prediction cash dosage pre-processed；
According to the transaction data in the history clear cycle of the atm device after pretreatment, geographical location, history clear week
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle in phase and the atm device mesh obtained in advance
The prediction cash dosage in clear cycle is marked, constructs the simulated environment model using supervised learning algorithm, wherein the simulation
Environmental model is for training data needed for generating described plus paper money planning optimization model.
4. atm device according to claim 3 adds paper money optimization method, which is characterized in that it is described according to pretreatment after
Transaction data, geographical location, the weather data in history clear cycle, paper money case in the history clear cycle of the atm device
Prediction in capacity, closing balance, remaining clear number of days, clear cycle and the atm device target clear cycle that obtains in advance
Cash dosage constructs the simulated environment model using supervised learning algorithm, comprising:
By changing the atm device closing balance and remaining clear number of days, the supervised learning algorithm and the simulation are utilized
Environmental model generates the simulated environment result under multiple simulated environments.
5. atm device according to claim 3 adds paper money optimization method, which is characterized in that the pretreatment includes: that data are clear
Reason, data integration, data regularization and data transformation.
6. atm device according to claim 4 adds paper money optimization method, which is characterized in that further include:
The training data under different simulated environments is generated according to multiple simulated environment models；
Described plus paper money planning optimization model is generated using nitrification enhancement according to the training data under different simulated environments, wherein
The nitrification enhancement includes: DQN algorithm, DDPG algorithm, IMPALA algorithm, A2C algorithm and GA3C algorithm.
7. atm device according to claim 1 adds paper money optimization method, which is characterized in that atm device add paper money operation include:
The atm device adds the paper money date and adds the paper money amount of money.
8. a kind of atm device adds paper money to optimize device characterized by comprising
Data acquisition first unit, the transaction data in current clear cycle, geographical location for obtaining the atm device,
Data of weather forecast, paper money tankage, closing balance and the remaining clear number of days in the geographical location；
Paper money is added to operate generation unit, for public holiday, described current clear in the target clear cycle according to the atm device
Transaction data, the geographical location in the machine period, paper money tankage, show the data of weather forecast in the target clear cycle
Golden remaining sum, remaining clear number of days based on preset simulated environment model and add paper money planning optimization model, generate the atm device
Paper money is added to operate.
9. atm device according to claim 8 adds paper money to optimize device, which is characterized in that the transaction data includes described
Pipelined data, pay day and the repayment date of atm device.
10. atm device according to claim 9 adds paper money to optimize device, which is characterized in that further include:
Data acquisition second unit, the transaction data in history clear cycle, geographical location for obtaining the atm device,
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle and acquisition in advance in history clear cycle
Atm device target clear cycle in prediction cash dosage；
Pretreatment unit, for transaction data, geographical location, the history clear in the history clear cycle to the atm device
Weather data, paper money tankage, closing balance, remaining clear number of days, clear cycle in period and the atm device obtained in advance
Prediction cash dosage in target clear cycle is pre-processed；
Environment geometric modeling unit, for the number of deals in the history clear cycle according to the atm device after pretreatment
According to weather data, paper money tankage, closing balance, the remaining clear number of days, clear cycle in, geographical location, history clear cycle
And the prediction cash dosage in the atm device target clear cycle obtained in advance, the mould is constructed using supervised learning algorithm
Quasi- environmental model, wherein the simulated environment model is for training data needed for generating described plus paper money planning optimization model.
11. atm device according to claim 10 adds paper money to optimize device, which is characterized in that environment geometric modeling unit tool
Body is used to utilize the supervised learning algorithm and the mould by changing the atm device closing balance and remaining clear number of days
Quasi- environmental model generates the simulated environment result under multiple simulated environments.
12. atm device according to claim 10 adds paper money to optimize device, which is characterized in that the pretreatment unit is specific
It is converted for data scrubbing, data integration, data regularization and data.
13. atm device according to claim 11 adds paper money to optimize device, which is characterized in that further include:
Training data generation unit, for generating the training data under different simulated environments according to multiple simulated environment models；
Add paper money model generation unit, described in generating according to the training data under different simulated environments using nitrification enhancement
Add paper money planning optimization model, wherein the nitrification enhancement includes: DQN algorithm, DDPG algorithm, IMPALA algorithm, A2C algorithm
And GA3C algorithm.
14. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that realize that any one of claim 1 to 7 atm device adds when the processor executes described program
The step of paper money optimization method.
15. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
The step of any one of claim 1 to 7 atm device adds paper money optimization method is realized when processor executes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910700928.3A CN110415462B (en) | 2019-07-31 | 2019-07-31 | ATM equipment banknote adding optimization method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910700928.3A CN110415462B (en) | 2019-07-31 | 2019-07-31 | ATM equipment banknote adding optimization method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110415462A true CN110415462A (en) | 2019-11-05 |
| CN110415462B CN110415462B (en) | 2021-09-03 |
Family
ID=68364511
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910700928.3A Active CN110415462B (en) | 2019-07-31 | 2019-07-31 | ATM equipment banknote adding optimization method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415462B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111667115A (en) * | 2020-06-05 | 2020-09-15 | 中国银行股份有限公司 | Cash processing method and device for self-service cash recycling machine |
| CN111695725A (en) * | 2020-06-02 | 2020-09-22 | 中国工商银行股份有限公司 | ATM (automatic teller machine) money adding method and device based on heuristic search |
| CN111708768A (en) * | 2020-05-20 | 2020-09-25 | 广州中智融通金融科技有限公司 | Cash dispatching method, system and storage medium for automatic teller machine |
| CN111986406A (en) * | 2020-09-03 | 2020-11-24 | 中国银行股份有限公司 | Data processing method and device based on withdrawal transaction |
| CN112819182A (en) * | 2021-01-22 | 2021-05-18 | 中国工商银行股份有限公司 | Training method, device and equipment for banknote box state prediction model |
| CN113593147A (en) * | 2021-08-03 | 2021-11-02 | 中国银行股份有限公司 | ATM balance display method, device, server, medium and product |
| EP3945507A1 (en) * | 2020-07-28 | 2022-02-02 | Planfocus software GmbH | Cash level optimization with autocalibration |
| CN114037169A (en) * | 2021-11-17 | 2022-02-11 | 中国农业银行股份有限公司深圳市分行 | Bank note distribution method and system for ATM equipment and storage medium |
| CN114358448A (en) * | 2022-03-21 | 2022-04-15 | 中国工商银行股份有限公司 | Driving route planning method and device |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104077631A (en) * | 2014-06-10 | 2014-10-01 | 通邮(中国)科技有限公司 | Cash management system of automatic teller machine |
| CN105447994A (en) * | 2014-08-27 | 2016-03-30 | 日立金融设备***（深圳）有限公司 | Cash bill case management system and cash bill case management method |
| CN107294993A (en) * | 2017-07-05 | 2017-10-24 | 重庆邮电大学 | A kind of WEB abnormal flow monitoring methods based on integrated study |
| CN107393212A (en) * | 2017-06-23 | 2017-11-24 | 彼易科技有限公司 | Multiple Currencies service system and method based on machine learning and distributed structure/architecture |
| WO2017222168A1 (en) * | 2016-06-20 | 2017-12-28 | 비씨카드(주) | Method for providing transaction data classification service for vat refund, and service providing server for performing same |
| CN108230547A (en) * | 2018-01-29 | 2018-06-29 | 深圳怡化电脑股份有限公司 | Add paper money management method plus paper money managing device and electronic equipment |
| CN108462517A (en) * | 2018-03-06 | 2018-08-28 | 东南大学 | A kind of MIMO link self-adaption transmission methods based on machine learning |
| WO2019039077A1 (en) * | 2017-08-23 | 2019-02-28 | 日本電気株式会社 | Cash demand forecast system, cash demand forecast method, and cash demand forecast program |
| CN109615760A (en) * | 2018-12-17 | 2019-04-12 | 山东省城市商业银行合作联盟有限公司 | Intelligently add the method and system of clear paper money to ATM in bank based on machine learning |
| CN109816530A (en) * | 2019-02-01 | 2019-05-28 | 中山大学 | A kind of financial trade method based on deeply study A3C algorithm |
| CN109829818A (en) * | 2019-02-03 | 2019-05-31 | 中国银行股份有限公司 | Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing |
| CN109840637A (en) * | 2019-02-20 | 2019-06-04 | 中国银行股份有限公司 | Cash demand amount prediction technique, device and equipment |
| CN110349356A (en) * | 2019-07-18 | 2019-10-18 | 中国银行股份有限公司 | Clear adds the monitoring system, monitoring method and monitoring device of paper money |
-
2019
- 2019-07-31 CN CN201910700928.3A patent/CN110415462B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104077631A (en) * | 2014-06-10 | 2014-10-01 | 通邮(中国)科技有限公司 | Cash management system of automatic teller machine |
| CN105447994A (en) * | 2014-08-27 | 2016-03-30 | 日立金融设备***（深圳）有限公司 | Cash bill case management system and cash bill case management method |
| WO2017222168A1 (en) * | 2016-06-20 | 2017-12-28 | 비씨카드(주) | Method for providing transaction data classification service for vat refund, and service providing server for performing same |
| CN107393212A (en) * | 2017-06-23 | 2017-11-24 | 彼易科技有限公司 | Multiple Currencies service system and method based on machine learning and distributed structure/architecture |
| CN107294993A (en) * | 2017-07-05 | 2017-10-24 | 重庆邮电大学 | A kind of WEB abnormal flow monitoring methods based on integrated study |
| WO2019039077A1 (en) * | 2017-08-23 | 2019-02-28 | 日本電気株式会社 | Cash demand forecast system, cash demand forecast method, and cash demand forecast program |
| CN108230547A (en) * | 2018-01-29 | 2018-06-29 | 深圳怡化电脑股份有限公司 | Add paper money management method plus paper money managing device and electronic equipment |
| CN108462517A (en) * | 2018-03-06 | 2018-08-28 | 东南大学 | A kind of MIMO link self-adaption transmission methods based on machine learning |
| CN109615760A (en) * | 2018-12-17 | 2019-04-12 | 山东省城市商业银行合作联盟有限公司 | Intelligently add the method and system of clear paper money to ATM in bank based on machine learning |
| CN109816530A (en) * | 2019-02-01 | 2019-05-28 | 中山大学 | A kind of financial trade method based on deeply study A3C algorithm |
| CN109829818A (en) * | 2019-02-03 | 2019-05-31 | 中国银行股份有限公司 | Cash demand amount prediction technique, device, electronic equipment and readable storage medium storing program for executing |
| CN109840637A (en) * | 2019-02-20 | 2019-06-04 | 中国银行股份有限公司 | Cash demand amount prediction technique, device and equipment |
| CN110349356A (en) * | 2019-07-18 | 2019-10-18 | 中国银行股份有限公司 | Clear adds the monitoring system, monitoring method and monitoring device of paper money |
Non-Patent Citations (2)
| Title |
|---|
| 徐呈康: "《银行ATM加钞路径及金额管理优化的研究与实现》", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 李德毅，于剑: "《人工智能导论》", 31 August 2018 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111708768A (en) * | 2020-05-20 | 2020-09-25 | 广州中智融通金融科技有限公司 | Cash dispatching method, system and storage medium for automatic teller machine |
| CN111695725A (en) * | 2020-06-02 | 2020-09-22 | 中国工商银行股份有限公司 | ATM (automatic teller machine) money adding method and device based on heuristic search |
| CN111667115A (en) * | 2020-06-05 | 2020-09-15 | 中国银行股份有限公司 | Cash processing method and device for self-service cash recycling machine |
| EP3945507A1 (en) * | 2020-07-28 | 2022-02-02 | Planfocus software GmbH | Cash level optimization with autocalibration |
| CN111986406A (en) * | 2020-09-03 | 2020-11-24 | 中国银行股份有限公司 | Data processing method and device based on withdrawal transaction |
| CN111986406B (en) * | 2020-09-03 | 2022-02-15 | 中国银行股份有限公司 | Data processing method and device based on withdrawal transaction |
| CN112819182A (en) * | 2021-01-22 | 2021-05-18 | 中国工商银行股份有限公司 | Training method, device and equipment for banknote box state prediction model |
| CN112819182B (en) * | 2021-01-22 | 2024-05-10 | 中国工商银行股份有限公司 | Training method, device and equipment for banknote box state prediction model |
| CN113593147A (en) * | 2021-08-03 | 2021-11-02 | 中国银行股份有限公司 | ATM balance display method, device, server, medium and product |
| CN114037169A (en) * | 2021-11-17 | 2022-02-11 | 中国农业银行股份有限公司深圳市分行 | Bank note distribution method and system for ATM equipment and storage medium |
| CN114358448A (en) * | 2022-03-21 | 2022-04-15 | 中国工商银行股份有限公司 | Driving route planning method and device |
| CN114358448B (en) * | 2022-03-21 | 2022-05-24 | 中国工商银行股份有限公司 | Driving route planning method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110415462B (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110415462A (en) | Atm device adds paper money optimization method and device | |
| Zheng et al. | The ai economist: Improving equality and productivity with ai-driven tax policies | |
| Zhu et al. | A mean-field Markov decision process model for spatial-temporal subsidies in ride-sourcing markets | |
| Wilson | Land-use/transport interaction models: Past and future | |
| Krafta | Modelling intraurban configurational development | |
| Zhou et al. | Agent-based simulation of electricity markets: a survey of tools | |
| Nooteboom | Diffusion, uncertainty and firm size | |
| CN107241213A (en) | A kind of web service composition method learnt based on deeply | |
| Rushton | The Roepke lecture in economic geography location theory, location-allocation models, and service development planning in the third world | |
| Zhang et al. | Mega-event marketing and urban growth coalitions: A case study of Nanjing Olympic New Town | |
| Jan et al. | The effect of clusters on the development of the software industry in Dalian, China | |
| Dotoli et al. | A Nash equilibrium simulation model for the competitiveness evaluation of the auction based day ahead electricity market | |
| CN115062926A (en) | Congestion relieving strategy determination method and device based on reinforcement learning and digital twinning | |
| Krishnamurty | Macroeconometric models for India: Past, present and prospects | |
| García et al. | Reinforcement learning for decision-making in a business simulator | |
| CN110135626A (en) | Credit management method and device, electronic equipment, storage medium | |
| Ribeiro et al. | Intelligent remuneration and tariffs for virtual power players | |
| CN105279978B (en) | Intersection traffic signal control method and equipment | |
| Zhang et al. | The economics of road network ownership: an agent-based approach | |
| Jamali et al. | Towards agent-based simulation of the parallel trading market of pharmaceuticals | |
| Prause et al. | Market model benchmark suite for machine learning techniques | |
| Lee | An integrated model for planning development projects using ACO and construction simulation | |
| Yoosefdoost et al. | Water resource management aided by game theory | |
| Perez-Trejo et al. | An exploration of dynamical systems modelling as a decision tool for environmental policy | |
| Nardone | Agent-based computational economics and industrial organization theory |
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 |