CN111783487B - Fault early warning method and device for card reader equipment - Google Patents

Fault early warning method and device for card reader equipment Download PDF

Info

Publication number
CN111783487B
CN111783487B CN202010585023.9A CN202010585023A CN111783487B CN 111783487 B CN111783487 B CN 111783487B CN 202010585023 A CN202010585023 A CN 202010585023A CN 111783487 B CN111783487 B CN 111783487B
Authority
CN
China
Prior art keywords
reading module
card reading
data
prediction
training
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
Application number
CN202010585023.9A
Other languages
Chinese (zh)
Other versions
CN111783487A (en
Inventor
高伟
郑广斌
钟春彬
杨洁琼
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Industrial and Commercial Bank of China Ltd ICBC
Original Assignee
Industrial and Commercial Bank of China Ltd ICBC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Industrial and Commercial Bank of China Ltd ICBC filed Critical Industrial and Commercial Bank of China Ltd ICBC
Priority to CN202010585023.9A priority Critical patent/CN111783487B/en
Publication of CN111783487A publication Critical patent/CN111783487A/en
Application granted granted Critical
Publication of CN111783487B publication Critical patent/CN111783487B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
    • G06K7/0095—Testing the sensing arrangement, e.g. testing if a magnetic card reader, bar code reader, RFID interrogator or smart card reader functions properly
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/18—Status alarms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/20—Ensemble learning
    • 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
    • G06N3/044—Recurrent networks, e.g. Hopfield 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
    • G06N3/04—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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound

Abstract

The invention provides a fault early warning method and device of card reader equipment, wherein the method comprises the following steps: acquiring prediction data of each card reading module of the card reader device; performing feature processing on the prediction data of each card reading module to obtain prediction feature data of each card reading module; obtaining the predicted use times of each card reading module according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module; and if the predicted accumulated use times of the card reading module are larger than the corresponding use threshold value, sending out fault early warning information. The device is used for executing the method. The fault early warning method and device for the card reader device provided by the embodiment of the invention improve the reliability of the card reader device.

Description

Fault early warning method and device for card reader equipment
Technical Field
The invention relates to the technical field of equipment maintenance, in particular to a fault early warning method and device for card reader equipment.
Background
The card reader device can be configured on the counter of a banking website and used for reading information of a bank card and an identity card.
In the prior art, the maintenance mode of the card reader equipment used by the counter of the banking website is fault maintenance, namely, the card reader equipment is maintained when the card reader equipment fails, but the card reader equipment is maintained when the card reader equipment fails, the use of the card reader equipment is influenced, and the replacement and maintenance of the card reader equipment also require time, so that the service cannot be provided for a customer within a certain time, the handling of customer business is influenced, and the customer service experience of the banking website is influenced.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a fault early warning method and device of card reader equipment, which can at least partially solve the problems in the prior art.
In one aspect, the present invention provides a fault early warning method for card reader equipment, including:
acquiring prediction data of each card reading module of the card reader device;
performing feature processing on the prediction data of each card reading module to obtain prediction feature data of each card reading module;
obtaining the predicted use times of each card reading module according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module;
Obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module;
and if the predicted accumulated use times of the card reading module are larger than the corresponding use threshold value, sending out fault early warning information.
In another aspect, the present invention provides a fault early warning apparatus for a card reader device, including:
the prediction data acquisition unit is used for acquiring the prediction data of each card reading module of the card reader device;
the first feature processing unit is used for carrying out feature processing on the prediction data of each card reading module to obtain the prediction feature data of each card reading module;
the prediction unit is used for obtaining the predicted use times of each card reading module according to the prediction characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module;
the obtaining unit is used for obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module;
And the judging unit is used for sending out fault early warning information after judging that the predicted accumulated use times of the card reading module are larger than the corresponding use threshold.
In yet another aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the steps of the fault early warning method of the card reader device of any of the above embodiments when the program is executed.
In yet another aspect, the present invention provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the fault pre-warning method of a card reader device according to any of the embodiments described above.
According to the method and the device for pre-warning faults of the card reader device, the prediction data of each card reading module of the card reader device can be obtained, the prediction data of each card reading module is subjected to feature processing to obtain the prediction feature data of each card reading module, the prediction use times of each card reading module are obtained according to the prediction feature data of each card reading module and the use times prediction model of each card reading module, the prediction accumulated use times of each card reading module are obtained according to the prediction use times and the current accumulated use times of each card reading module, fault pre-warning information is sent after judging that the prediction accumulated use times of the card reading module are larger than the corresponding use threshold, fault pre-warning can be carried out before the service life of the card reading module is expired, so that actions are taken in advance to maintain or replace the card reading module, active maintenance of the card reader device is achieved, and reliability of the card reader device is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
fig. 1 is a flowchart of a fault early warning method of a card reader device according to an embodiment of the present invention.
Fig. 2 is a flowchart of a fault early warning method of a card reader device according to another embodiment of the present invention.
Fig. 3 is a flowchart of a fault early warning method of a card reader device according to another embodiment of the present invention.
Fig. 4 is a flowchart of a fault early warning method of a card reader device according to still another embodiment of the present invention.
Fig. 5 is a schematic structural diagram of a fault early warning device of a card reader device according to an embodiment of the present invention.
Fig. 6 is a schematic structural diagram of a fault early warning device of a card reader device according to another embodiment of the present invention.
Fig. 7 is a schematic structural diagram of a fault early warning device of a card reader device according to another embodiment of the present invention.
Fig. 8 is a schematic structural diagram of a fault early warning device of a card reader device according to still another embodiment of the present application.
Fig. 9 is a schematic physical structure of an electronic device according to an embodiment of the application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the embodiments of the present application will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present application and their descriptions herein are for the purpose of explaining the present application, but are not to be construed as limiting the application. It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be arbitrarily combined with each other.
Fig. 1 is a flow chart of a fault early warning method of a card reader device according to an embodiment of the present application, as shown in fig. 1, where the fault early warning method of a card reader device according to an embodiment of the present application includes:
s101, acquiring prediction data of each card reading module of card reader equipment;
specifically, the prediction data of each card reading module of the card reader device may be obtained by preprocessing the historical record data of the past first preset time period of each card reading module of the card reader device, and the server may obtain the prediction data of each card reading module of the card reader device. The card reading module can be any one of a magnetic stripe type reading and writing module, a contact type IC card reading and writing module, a radio frequency IC card reading and writing module and an identity card reading module, and the card reader device comprises at least one card reading module. The history data of each card reading module may include data of the device itself, periodic data, payroll and social security data, holiday data, usage count data, contemporaneous usage count data, past usage count data, average usage count data, and ring ratio usage count data, etc., which are set according to actual needs, and the embodiment of the application is not limited. The first preset time period is, for example, 3 months, and is set according to actual needs, which is not limited in the embodiment of the present application. The execution main body of the fault early warning method of the card reader device provided by the embodiment of the application comprises, but is not limited to, a server.
For example, historical data for each card reading module of the card reader device may be obtained by a banking internal system or web crawler. The device itself data may include information about the location of the region to which it belongs, the network point to which it belongs, the number of card reader devices in the network point, the window of use (for the public or the individual), etc. The periodic data may include month, month date/month end, month date week, day of week, etc. The payroll and social security data may include information of whether payroll was done, whether payroll was done for the first N days, whether payroll was done for the next N days, whether social security was done for the first N days, whether social security was done for the next N days, and whether social security was done for the next N days. The holiday data includes information such as whether the holiday is present, whether several days are present in the first N days, and whether several days are present in the last N days. The usage count data may include information such as a maximum value of the number of usage times in the previous N days, a minimum value of the number of usage times in the previous N days, an average value of the number of usage times in the previous N days, a median of the number of usage times in the previous N days, a variance of the number of usage times in the previous N days, and the like. The contemporaneous usage frequency data may include information such as the number of usage times of the last week and the same day, the number of usage times of the last month and the same day, the number of usage times of the last year and the like. The past usage number data may include yesterday usage number, previous three-day usage number, previous four-day usage number, and the like. The average usage number data may include information such as average usage number of the last week, average usage number of the last month and the same day of the week, etc. The cycle ratio usage number data may include information such as the cycle ratio of the previous N days, zhou Huanbi of the previous N days, the cycle ratio of the previous N days, and the cycle ratio of the previous N days. In order to predict the predicted use times of the card reading module in the next day, the value range of N is greater than or equal to 1 and less than or equal to 7, and is set according to the actual situation, which is not limited in the embodiment of the present invention.
For example, preprocessing the history data of each card reading module may include correcting inconsistencies in the history data by filling in missing values, smoothing noise, identifying outliers, and the like. When filling the missing values, adopting an average value and a median to fill; when processing noise data and outliers, firstly determining the noise data through the upper edge and the lower edge of a box diagram, detecting the outliers through a clustering algorithm, marking the noise data and the outliers by combining specific service experience, and finally calculating average values of a plurality of periods before and after to repair the noise data and the outliers.
Preprocessing the history data of each card reading module may include integrating the history data. Because the data acquired by the system has various sources, the attribute representing the same concept may have different names or units in different data sources, which may lead to data inconsistency and redundancy.
Preprocessing the history data for each card reading module may include specifying the history data for each card reading module. The simplified representation of the data can be obtained through the reduction technology, the occupied space of the simplified data can be reduced, but the nearly same analysis result can be generated, and the data processing efficiency of the fault early warning method of the card reader device can be improved.
Preprocessing the history data of each card reading module may include performing data transformation on the history data of each card reading module, and the history data of each card reading module is more suitable for performing data mining through the data transformation. For example, the transformation of the geographical position information classifies the geographical position information, and the same category uses the same digital representation, so that the text data is transformed into discrete numerical data.
S102, performing feature processing on the prediction data of each card reading module to obtain the prediction feature data of each card reading module;
specifically, after obtaining the prediction data of each card reading module, the server may perform feature processing on the prediction data of each card reading module to obtain the prediction feature data of each card reading module.
For example, feature processing of the prediction data for each card reading module may include feature construction. The feature construction is used for constructing different types of features from the prediction data of each card reading module, the device self feature can be constructed based on the device self data, the periodic feature can be constructed based on the periodic data, the payroll social security feature can be constructed based on the payroll and social security data, the holiday feature can be constructed based on the holiday data, the use number statistics feature can be constructed based on the use number statistics data, the contemporaneous use feature can be constructed based on the contemporaneous use number data, the past use feature can be constructed based on the past use number data, the average use feature can be constructed based on the average use number data, and the ring ratio use feature can be constructed based on the ring ratio use number data.
S103, obtaining the predicted use times of each card reading module according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module;
specifically, after obtaining the prediction feature data of each card reading module, the server may input the prediction feature data of each card reading module into the usage number prediction model of each card reading module, and output the predicted usage number of each card reading module through the processing of the usage number prediction model of each card reading module. The use number prediction model of each card reading module comprises a set number of predictor models, and the set number of predictor models are obtained according to the use number training data and the historical use number training of each card reading module. The set number is set according to actual situations, and the embodiment of the invention is not limited.
S104, obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module;
Specifically, after obtaining the predicted usage times of each card reading module, the server may also obtain the current accumulated usage times of each card reading module, where each card reading module is recorded when in use, and may count to obtain the current accumulated usage times of each card reading module. For each card reading module, the server calculates the sum of the predicted use times and the current accumulated use times of the card reading module, and takes the summed result as the predicted accumulated use times of the card reading module. The current accumulated use times of the card reading module refer to total use times before the card reading module starts to be put into use and predicts the use times.
And S105, if judging that the predicted accumulated use times of the card reading module are greater than the corresponding use threshold, sending out fault early warning information.
Specifically, after obtaining the predicted and accumulated use times of each card reading module, the server compares the predicted and accumulated use times of the card reading module with the use threshold corresponding to the card reading module for each card reading module, and if the predicted and accumulated use times of the card reading module are greater than the use threshold corresponding to the card reading module, the server indicates that the card reading module is about to fail, and then the server can send out failure early warning information so as to facilitate operation and maintenance preparation work of staff. If the predicted accumulated use times of the card reading module are smaller than or equal to the use threshold corresponding to the card reading module, the card reading module can still be used without maintenance. The usage threshold value corresponding to each card reading module is set according to the actual situation, and the embodiment of the invention is not limited. It can be understood that the predicted accumulated usage times of any card reading module of the card reader device are greater than the corresponding usage threshold value, fault early warning information can be sent out, and the card reading module needing maintenance can be indicated in the fault early warning information.
For example, the card reading module is a magnetic stripe type read-write module, the card reading module is expected to be able to be swiped 2000 times, that is, the usage threshold value corresponding to the magnetic stripe type read-write module is 2000 times, if the server obtains the predicted usage frequency of the magnetic stripe type read-write module for 300 times in the next day and obtains the current accumulated usage frequency of the magnetic stripe type read-write module for 1800 times, the server calculates the predicted accumulated usage frequency of the magnetic stripe type read-write module for 1800+300=2100 times, and because the predicted accumulated usage frequency 2100 of the magnetic stripe type read-write module is greater than the usage threshold value 2000 corresponding to the magnetic stripe type read-write module, the server sends fault early warning information, for example, informs a worker in a mode of mobile phone short messages, and the magnetic stripe type read-write module can reach the usage threshold value in the next day and needs to be replaced.
According to the method for pre-warning faults of the card reader device, the pre-warning data of each card reading module of the card reader device can be obtained, the pre-warning data of each card reading module is subjected to feature processing, the pre-warning feature data of each card reading module is obtained, the pre-warning use times of each card reading module are obtained according to the pre-warning feature data of each card reading module and the use times of each card reading module, the pre-warning use times of each card reading module are obtained according to the pre-warning use times of each card reading module and the current accumulated use times, after judging that the pre-warning accumulated use times of the card reading module are larger than corresponding use thresholds, fault pre-warning information is sent out, and fault pre-warning can be carried out before the service life of the card reading module is expired, so that actions are taken in advance to maintain or replace the card reading module, active maintenance of the card reader device is achieved, and reliability of the card reader device is improved. In addition, interruption to business handling due to failure of the card reader device can be reduced, and influence on customer service is reduced.
Fig. 2 is a flow chart of a fault early warning method of a card reader device according to another embodiment of the present invention, as shown in fig. 2, further, based on the foregoing embodiments, the obtaining, according to the prediction feature data of each card reading module and the usage number prediction model of each card reading module, the predicted usage number of each card reading module includes:
s1031, obtaining the set number of times of use of the card reading module according to the predicted characteristic data of the card reading module and the set number of predicted sub-models;
specifically, the server inputs the prediction feature data of the card reading module to each prediction sub-model, and can output the use times of the card reading module under each prediction sub-model, wherein the number of the prediction sub-models is set, and then the server can obtain the set number of the use times of the card reading module.
S1032, obtaining the predicted use times of the card reading module by setting a number of use times of the card reading module and weights corresponding to the use times; the weight corresponding to each use frequency of the card reading module is obtained in advance.
Specifically, after obtaining the set number of usage times of the card reading module, the server calculates the predicted usage times of the card reading module according to the set number of usage times of the card reading module and weights corresponding to the usage times. The weight corresponding to each usage number is obtained in advance, for example, preset or dynamically generated.
For example, the card reading module is an identity card reading module, and the server obtains n identity card reading modulesThe number of times of use can be according to the formulaCalculating to obtain the predicted use times P of the identity card reading module, wherein Q i The ith use number k of the identity card reading module i Is Q i And the corresponding weight, i is a positive integer and i is smaller than n.
The weight corresponding to each usage number of each card reading module may be pre-generated according to the number of predictor models and the weight vector generation algorithm. The first step, given N and H, N is the number of predictor models, namely the set number, 1/H represents the granularity of weight change, the set M= {1, 1., 1}, and the set M contains H1 s in total; secondly, 1 in the set M is divided into N groups by using an interpolation method, and the common set can be obtained by the arrangement and combination thoughtA seed distribution mode; thirdly, adding up 1 of each group in each distribution mode and dividing by H to obtain the common +.>A set of evenly distributed weight vectors.
For example, in the first step, let n=2 and h=3, i.e. the number of predictor models is 2 and the granularity of the weight change is 3. Secondly, according to the insertion method in the arrangement idea, 4 distribution modes are { { zero 1, three 1}, {1, two 1}, { two 1, one 1}, { three 1, zero 1 }. Thirdly, adding 1 in each allocation mode to obtain { {0,3}, {1,2}, {2,1}, {3,0}, and dividing H=3 to obtain weight vector sets as follows: {0,1}, {1/3,2/3}, {2/3,1/3}, {1,0}, for four sets of weights.
After the vector set of the weights is obtained, the effect of each predictor model on each group of weight vectors in the weight vector set can be evaluated according to the verification set of the card reading module and the historical use times corresponding to the verification set, and the optimal group of weights is selected. Firstly, according to the verification set and each predictor model, obtaining the corresponding use times of each predictor model; then, according to the corresponding use times of each predictor model and each group of weights, calculating the combined use times of the card reading module under each group of weights; and finally, calculating the RMSE value of the combined use times and the historical use times corresponding to each group of weights by using an RMSE (standard error) method, namely, calculating the sum of the root mean square error of the combined use times and the historical use times of the card reading module in the verification set for each group of weights, wherein the group of weights with the minimum RMSE value is the optimal weight vector. The process of obtaining the verification set of the card reading module is described below.
Fig. 3 is a flowchart of a fault early warning method of a card reader device according to another embodiment of the present invention, as shown in fig. 3, further, based on the foregoing embodiments, obtaining a usage number prediction model of each card reading module according to usage number training data and historical usage number training of each card reading module includes:
S301, acquiring the training data of the use times of the card reading module and the historical use times;
specifically, the training data of the number of times of use of each card reading module can be obtained by preprocessing the history data of the past second preset time period of each card reading module of the card reader device. The historical usage times can be the usage times of each card reading module in the past within a second preset time period, and the historical usage times are set according to actual needs, and the embodiment of the invention is not limited. The server can acquire the training data of the use times and the historical use times of the card reading module. The second preset time period is, for example, 3 years, and is set according to actual needs, which is not limited in the embodiment of the present invention. It can be understood that if the historical usage number is the usage number of each day, the usage number prediction model of the card reading module obtained by training predicts the usage number of the card reading module in the next day; if the historical usage number is the usage number of every three days, the usage number prediction model of the card reading module obtained through training predicts the future usage number of the card reading module in three days.
S302, obtaining the use frequency characteristic data of the card reading module according to the use frequency training data of the card reading module;
specifically, after obtaining the usage training data of the card reading module, the server may perform feature processing on the usage training data of the card reading module to obtain usage feature data of the card reading module.
For example, different types of features are constructed from the usage count training data of the card reading module, the device itself feature can be constructed based on the device itself data, the periodic feature can be constructed based on the periodic data, the payroll social security feature can be constructed based on the payroll and social security data, the holiday feature can be constructed based on the holiday data, the usage count statistics feature can be constructed based on the usage count data, the contemporaneous usage feature can be constructed based on the contemporaneous usage count data, the past usage feature can be constructed based on the past usage count data, the average usage feature can be constructed based on the average usage count data, and the ring ratio usage feature can be constructed based on the ring ratio usage count data.
S303, dividing the usage frequency characteristic data of the card reading module into a training set and a verification set;
Specifically, after obtaining the usage frequency characteristic data of the card reading module, the server may divide the usage frequency characteristic data of the card reading module into a training set and a verification set, where the training set is used for model training, and the verification set is used for model verification.
For example, the training set is 80% of the usage frequency characteristic data of the card reading module, the verification set is 20% of the usage frequency characteristic data of the card reading module, and the training set and the verification set are not overlapped.
S304, training to obtain a preset number of to-be-determined predictive sub-models according to the training set, the historical use times corresponding to the training set and the preset number of preset models; wherein the preset number is greater than or equal to the set number;
specifically, the server performs model training on each preset model according to the training set and the historical use times corresponding to the training set, and obtains each to-be-determined prediction sub-model. The number of the undetermined predictive sub-models is preset, and the preset undetermined predictive sub-models can be obtained. Wherein the preset number is greater than or equal to the set number. The preset number is set according to actual needs, and the embodiment of the invention is not limited. The preset model includes, but is not limited to, a support vector machine regression algorithm, a K nearest neighbor regression algorithm, a random forest regression algorithm, a GBDT regression algorithm, an xgboost regression algorithm, a Long Short-Term Memory (LSTM) algorithm, etc., and is set according to actual needs, and the embodiment of the present invention is not limited. The LSTM algorithm is a neural network algorithm that can train a model by back-propagation. It can be understood that, in the process of training each preset model, the super-parameter automatic tuning can be performed until the parameters that make the model prediction effect globally optimal or locally optimal are obtained, in the process of super-parameter tuning, the super-parameter that makes the model prediction effect optimal is dynamically found based on the preset parameter value range and the value change step length, and the super-parameter search algorithm used includes, but is not limited to, grid search (gridSearchCV), random search (RandomazedSearchCV), and the like, and is set according to actual needs.
S305, verifying the preset number of the undetermined sub-models according to the verification set and the historical use times corresponding to the verification set to obtain an R square value of each undetermined sub-model;
specifically, after obtaining the preset number of pending predictor models, the server inputs the verification set into the pending predictor models for each pending predictor model, and outputs the estimated use times corresponding to each feature data in the verification set. The server may calculate an R-squared (R-squared) value of the pending predictor model according to the estimated number of uses corresponding to each feature data in the validation set and the historical number of uses corresponding to each feature data in the validation set. The server may obtain R square values for the predetermined number of pending predictor models. The larger the R square value is, the better the use times prediction effect of the undetermined predictor model is.
S306, according to the R square value of each undetermined predictor model, and selecting the set number of predictive sub-models from the preset number of predictive sub-models.
Specifically, after obtaining the R square values of the preset number of pending predictor models, the server may sort the R square values of the preset number of pending predictor models according to the magnitude of the R square values, and if the absolute value of the difference value of the R square values of all adjacent two pending predictor models is smaller than the preset value, it is indicated that the R square values of the preset number of pending predictor models are not greatly different, and the preset number of pending predictor models may be taken as the set number of predictor models. The preset value is set according to practical experience, and the embodiment of the invention is not limited.
If the absolute value of the difference value of the R square values of the two adjacent to-be-determined predictive sub-models is larger than or equal to the preset value, the R square values of the preset number of to-be-determined predictive sub-models are larger, the to-be-determined predictive sub-model with the largest R square value in the set number can be taken as the sales predictive sub-model with the set number larger than or equal to 4.
And selecting a to-be-determined predictive sub-model with the R square value meeting the requirement from a preset number of to-be-determined predictive sub-models as a predictive sub-model for use frequency prediction, so that the accuracy of use frequency prediction of the card reading module can be improved.
Fig. 4 is a flowchart of a fault early warning method of a card reader device according to still another embodiment of the present invention, as shown in fig. 4, further, based on the foregoing embodiments, the obtaining the usage number feature data of the card reading module according to the usage number training data of the card reading module includes:
s3021, performing feature construction on the training data of the use times of the card reading module to obtain multi-class training feature data;
specifically, after obtaining the training data of the usage times of the card reading module, the server may perform feature construction on the training data of the usage times of the card reading module to obtain multiple types of training feature data. The multi-class training feature data may include a device self feature, a period feature, a payroll social security feature, a holiday feature, a use number statistics feature, a contemporaneous use feature, a past use feature, an average use feature, a ring ratio use feature, and the like, which are set according to actual needs, and the embodiment of the invention is not limited.
S3022, selecting the use times characteristic data of the card reading module from the multiple training characteristic data according to a characteristic selection algorithm.
Specifically, after the server obtains the multiple types of training feature data, a feature selection algorithm may be used to select the usage frequency feature data of the card reading module from the multiple types of training feature data. The feature selection algorithm includes but is not limited to an algorithm such as directional search, optimal priority search, sequence forward selection, sequence backward selection, sequence floating selection, and the like, and is set according to practical situations, and the embodiment of the invention is not limited. The characteristic selection algorithm is used for selecting the use times characteristic data of the card reading module from the multiple types of training characteristic data, so that the prediction of the use times of the card reading module can be more accurately realized. It is understood that the usage count feature data includes at least usage count statistics.
It can be understood that after the fault early warning method of the card reader device provided by the embodiment of the invention is put into use, the actual use times of each card reading module can be collected, when a certain number of the card reading modules are collected, the R square value of the use times prediction model of the card reading modules can be calculated according to the actual use times and the predicted use times of the card reading modules, and the use times prediction model of the card reading modules is evaluated through the R square value. And when the R square value of the usage frequency prediction model of the card reading module is smaller than a set value, retraining to obtain the usage frequency prediction model of the card reading module. Wherein the certain number and the set value are set according to practical experience, and the embodiment of the invention is not limited.
Fig. 5 is a schematic structural diagram of a fault early warning device of a card reader device according to an embodiment of the present invention, where, as shown in fig. 5, the fault early warning device of a card reader device according to an embodiment of the present invention includes a predicted data obtaining unit 501, a first feature processing unit 502, a predicting unit 503, an obtaining unit 504, and a judging unit 505, where:
the predicted data obtaining unit 501 is configured to obtain predicted data of each card reading module of the card reader device; the first feature processing unit 502 is configured to perform feature processing on the predicted data of each card reading module, so as to obtain predicted feature data of each card reading module; the prediction unit 503 is configured to obtain a predicted usage number of each card reading module according to the prediction feature data of each card reading module and the usage number prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module; the obtaining unit 504 is configured to obtain a predicted cumulative usage number of each card reading module according to the predicted usage number and the current cumulative usage number of each card reading module; the judging unit 505 is configured to send out fault early warning information after judging that the predicted accumulated usage times of the card reading module is greater than the corresponding usage threshold.
Specifically, by preprocessing the history data of the past first preset time period of each card reading module of the card reader device, the prediction data of each card reading module may be obtained, and the obtaining unit 501 may obtain the prediction data of each card reading module of the card reader device. The card reading module can be any one of a magnetic stripe type reading and writing module, a contact type IC card reading and writing module, a radio frequency IC card reading and writing module and an identity card reading module, and the card reader device comprises at least one card reading module. The history data of each card reading module may include data of the device itself, periodic data, payroll and social security data, holiday data, usage count data, contemporaneous usage count data, past usage count data, average usage count data, and ring ratio usage count data, etc., which are set according to actual needs, and the embodiment of the invention is not limited. The first preset time period is, for example, 3 months, and is set according to actual needs, which is not limited in the embodiment of the present invention.
After obtaining the prediction data of each card reading module, the first feature processing unit 502 may perform feature processing on the prediction data of each card reading module to obtain the prediction feature data of each card reading module.
After obtaining the prediction feature data of each card reading module, the prediction unit 503 may input the prediction feature data of each card reading module into the usage number prediction model of each card reading module, and output the predicted usage number of each card reading module through the processing of the usage number prediction model of each card reading module. The use number prediction model of each card reading module comprises a set number of predictor models, and the set number of predictor models are obtained according to the use number training data and the historical use number training of each card reading module. The set number is set according to actual situations, and the embodiment of the invention is not limited.
After obtaining the predicted usage times of each card reading module, the obtaining unit 504 may further obtain the current accumulated usage times of each card reading module, where each card reading module is recorded during usage, and may statistically obtain the current accumulated usage times of each card reading module. For each card reading module, the obtaining unit 504 calculates the sum of the predicted usage times and the current accumulated usage times of the card reading module, and uses the result of the sum as the predicted accumulated usage times of the card reading module. The current accumulated use times of the card reading module refer to total use times before the card reading module starts to be put into use and predicts the use times.
After obtaining the predicted cumulative usage times of each card reading module, for each card reading module, the judging unit 505 compares the predicted cumulative usage times of the card reading module with the usage threshold corresponding to the card reading module, and if the predicted cumulative usage times of the card reading module are greater than the usage threshold corresponding to the card reading module, it indicates that the card reading module is about to fail, then the server may send failure early warning information so as to facilitate operation and maintenance preparation for staff. If the predicted accumulated use times of the card reading module are smaller than or equal to the use threshold corresponding to the card reading module, the card reading module can still be used without maintenance. The usage threshold value corresponding to each card reading module is set according to the actual situation, and the embodiment of the invention is not limited. It can be understood that the predicted accumulated usage times of any card reading module of the card reader device are greater than the corresponding usage threshold value, fault early warning information can be sent out, and the card reading module needing maintenance can be indicated in the fault early warning information.
The fault early warning device of the card reader device provided by the embodiment of the invention can acquire the prediction data of each card reading module of the card reader device, perform feature processing on the prediction data of each card reading module to acquire the prediction feature data of each card reading module, acquire the prediction use times of each card reading module according to the prediction feature data of each card reading module and the use times prediction model of each card reading module, acquire the prediction accumulated use times of each card reading module according to the prediction use times and the current accumulated use times of each card reading module, send out fault early warning information after judging that the prediction accumulated use times of the card reading module are larger than the corresponding use threshold, and perform fault early warning before the service life of the card reading module is expired so as to take action in advance to maintain or replace the card reading module, thereby realizing active maintenance of the card reader device and improving the reliability of the card reader device. In addition, interruption to business handling due to failure of the card reader device can be reduced, and influence on customer service is reduced.
Fig. 6 is a schematic structural diagram of a fault early warning device of a card reader device according to another embodiment of the present invention, as shown in fig. 6, further, based on the above embodiments, a prediction unit 503 includes a prediction subunit 5031 and a calculation subunit 5032, where:
the predictor unit 5031 is configured to obtain a set number of usage times of the card reading module according to the predicted feature data of the card reading module and the set number of predictor models; the calculating subunit 5032 is configured to calculate a predicted usage number of the card reading module according to the set number of usage numbers of the card reading module and weights corresponding to the usage numbers; the weight corresponding to each use frequency of the card reading module is obtained in advance.
Specifically, the prediction sub-unit 5031 inputs the prediction feature data of the card reading module to each prediction sub-model, and may output the number of times the card reading module is used under each prediction sub-model, where there are a set number of prediction sub-models, and then the server may obtain the set number of times the card reading module is used.
After obtaining the set number of usage times of the card reading module, the calculating subunit 5032 calculates the predicted usage times of the card reading module according to the set number of usage times of the card reading module and the weights corresponding to the respective usage times. The weight corresponding to each usage number is obtained in advance, for example, preset or dynamically generated.
Fig. 7 is a schematic structural diagram of a fault early warning device of a card reader device according to another embodiment of the present invention, as shown in fig. 7, further, based on the foregoing embodiments, the fault early warning device of a card reader device according to the embodiment of the present invention further includes a training data obtaining unit 506, a second feature processing unit 507, a dividing unit 508, a training unit 509, a verification unit 510, and a selecting unit 511, where:
the training data obtaining unit 506 is configured to obtain training data of the number of times of use and historical number of times of use of the card reading module; the second feature processing unit 507 is configured to obtain usage frequency feature data of the card reading module according to usage frequency training data of the card reading module; the dividing unit 508 is used for dividing the usage frequency characteristic data of the card reading module into a training set and a verification set; the training unit 509 is configured to perform training according to the training set, the historical usage times corresponding to the training set, and a preset number of preset models to obtain a preset number of to-be-determined predictor models; wherein the preset number is greater than or equal to the set number; the verification unit 510 is configured to verify the preset number of pending predictor models according to the verification set and the historical usage times corresponding to the verification set, to obtain an R square value of each pending predictor model; the selection unit 511 is arranged for based on the R-squared value of each of the pending predictor models, and selecting the set number of predictive sub-models from the preset number of predictive sub-models.
Specifically, the training data of the number of times of use of each card reading module can be obtained by preprocessing the history data of the past second preset time period of each card reading module of the card reader device. The historical usage times can be the usage times of each card reading module in the past within a second preset time period, and the historical usage times are set according to actual needs, and the embodiment of the invention is not limited. The training data obtaining unit 506 may obtain the training data of the number of uses of the card reading module and the historical number of uses. The second preset time period is, for example, 3 years, and is set according to actual needs, which is not limited in the embodiment of the present invention. It can be understood that if the historical usage number is the usage number of each day, the usage number prediction model of the card reading module obtained by training predicts the usage number of the card reading module in the next day; if the historical usage number is the usage number of every three days, the usage number prediction model of the card reading module obtained through training predicts the future usage number of the card reading module in three days.
After obtaining the usage training data of the card reading module, the second feature processing unit 507 may perform feature processing on the usage training data of the card reading module to obtain usage feature data of the card reading module.
After obtaining the usage frequency characteristic data of the card reading module, the dividing unit 508 may divide the usage frequency characteristic data of the card reading module into a training set for model training and a verification set for model verification.
The training unit 509 performs model training on each preset model according to the training set and the historical usage times corresponding to the training set, to obtain each to-be-determined prediction sub-model. The number of the undetermined predictive sub-models is preset, and the preset undetermined predictive sub-models can be obtained. Wherein the preset number is greater than or equal to the set number. The preset number is set according to actual needs, and the embodiment of the invention is not limited. The preset model includes, but is not limited to, a support vector machine regression algorithm, a K nearest neighbor regression algorithm, a random forest regression algorithm, a GBDT regression algorithm, an xgboost regression algorithm, an LSTM algorithm, and the like, and is set according to actual needs, and the embodiment of the invention is not limited. The LSTM algorithm is a neural network algorithm that can train a model by back-propagation. It can be understood that, in the process of training each preset model, the super-parameter automatic tuning can be performed until the parameters for enabling the model prediction effect to be globally optimal or locally optimal are obtained, in the process of super-parameter tuning, the super-parameter for enabling the model prediction effect to be optimal is dynamically searched based on the preset parameter value range and the value change step length, the used super-parameter search algorithm comprises, but is not limited to, gridSearchCV, randomizedSearchCV and the like, and is set according to actual needs, and the embodiment of the invention is not limited.
After the preset number of pending predictor models are obtained, for each pending predictor model, the verification unit 510 inputs the verification set into the pending predictor model, and outputs the estimated number of times of use corresponding to each feature data in the verification set. The verification unit 510 may calculate an R square (R-squared) value of the predetermined predictor model according to the estimated number of times of use corresponding to each feature data in the verification set and the historical number of times of use corresponding to each feature data in the verification set. The verification unit 510 may obtain R square values of the predetermined number of pending predictor models. The larger the R square value is, the better the use times prediction effect of the undetermined predictor model is.
After obtaining the R square values of the preset number of pending predictor models, the selecting unit 511 may sort the R square values of the preset number of pending predictor models according to the magnitude of the R square values, and if the absolute values of the differences of the R square values of all adjacent two pending predictor models are smaller than the preset value, it is indicated that the R square values of the preset number of pending predictor models are not greatly different, and may take the preset number of pending predictor models as the set number of predictor models. The preset value is set according to practical experience, and the embodiment of the invention is not limited.
If the absolute value of the difference value of the R square values of the two adjacent to-be-determined predictive sub-models is larger than or equal to the preset value, the R square values of the preset number of to-be-determined predictive sub-models are larger, the to-be-determined predictive sub-model with the largest R square value in the set number can be taken as the sales predictive sub-model with the set number larger than or equal to 4.
Fig. 8 is a schematic structural diagram of a fault early warning device of a card reader device according to still another embodiment of the present invention, as shown in fig. 8, further, based on the above embodiments, the second feature processing unit 507 includes a constructing subunit 5071 and a selecting subunit 5072, where:
the construction subunit 5071 is configured to perform feature construction on the training data of the usage times of the card reading module, so as to obtain multiple types of training feature data; the selecting subunit 5072 is configured to select the usage frequency feature data of the card reading module from the multiple types of training feature data according to a feature selection algorithm.
Specifically, after obtaining the training data of the number of uses of the card reading module, the construction subunit 5071 may perform feature construction on the training data of the number of uses of the card reading module, to obtain multiple types of training feature data. The multi-class training feature data may include a device self feature, a period feature, a payroll social security feature, a holiday feature, a use number statistics feature, a contemporaneous use feature, a past use feature, an average use feature, a ring ratio use feature, and the like, which are set according to actual needs, and the embodiment of the invention is not limited.
After obtaining the multiple types of training feature data, the selecting subunit 5072 may use a feature selection algorithm to select the usage number feature data of the card reading module from the multiple types of training feature data. The feature selection algorithm includes but is not limited to an algorithm such as directional search, optimal priority search, sequence forward selection, sequence backward selection, sequence floating selection, and the like, and is set according to practical situations, and the embodiment of the invention is not limited. The characteristic selection algorithm is used for selecting the use times characteristic data of the card reading module from the multiple types of training characteristic data, so that the prediction of the use times of the card reading module can be more accurately realized. It is understood that the usage count feature data includes at least usage count statistics.
The embodiment of the apparatus provided in the embodiment of the present invention may be specifically used to execute the processing flow of each method embodiment, and the functions thereof are not described herein again, and may refer to the detailed description of the method embodiments.
Fig. 9 is a schematic physical structure of an electronic device according to an embodiment of the present invention, as shown in fig. 9, the electronic device may include: processor 901, communication interface (Communications Interface) 902, memory 903 and communication bus 904, wherein processor 901, communication interface 902 and memory 903 communicate with each other via communication bus 904. The processor 901 may call logic instructions in the memory 903 to perform the following method: acquiring prediction data of each card reading module of the card reader device; performing feature processing on the prediction data of each card reading module to obtain prediction feature data of each card reading module; obtaining the predicted use times of each card reading module according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module; obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module; and if the predicted accumulated use times of the card reading module are larger than the corresponding use threshold value, sending out fault early warning information.
Further, the logic instructions in the memory 903 described above may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a stand alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The present embodiment discloses a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, are capable of performing the methods provided by the above-described method embodiments, for example comprising: acquiring prediction data of each card reading module of the card reader device; performing feature processing on the prediction data of each card reading module to obtain prediction feature data of each card reading module; obtaining the predicted use times of each card reading module according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module; obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module; and if the predicted accumulated use times of the card reading module are larger than the corresponding use threshold value, sending out fault early warning information.
The present embodiment provides a computer-readable storage medium storing a computer program that causes the computer to execute the methods provided by the above-described method embodiments, for example, including: acquiring prediction data of each card reading module of the card reader device; performing feature processing on the prediction data of each card reading module to obtain prediction feature data of each card reading module; obtaining the predicted use times of each card reading module according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module; obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module; and if the predicted accumulated use times of the card reading module are larger than the corresponding use threshold value, sending out fault early warning information.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In the description of the present specification, reference to the terms "one embodiment," "one particular embodiment," "some embodiments," "for example," "an example," "a particular example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.

Claims (10)

1. A method for fault early warning of a card reader device, comprising:
acquiring prediction data of each card reading module of the card reader device; the prediction data of each card reading module is obtained by preprocessing historical record data of a past first preset time period of each card reading module of the card reader device;
performing feature processing on the prediction data of each card reading module to obtain prediction feature data of each card reading module; the method comprises the steps that the feature processing of the prediction data of each card reading module comprises feature construction, wherein the feature construction is used for constructing different types of features from the prediction data of each card reading module;
obtaining the predicted use times of each card reading module in a future period according to the predicted characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module;
obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times of each card reading module in a future period of time; the predicted accumulated use times of the card reading module are equal to the sum of the predicted use times and the current accumulated use times in a future period of time of the card reading module;
And if the predicted accumulated use times of the card reading module are larger than the corresponding use threshold value, sending out fault early warning information.
2. The method of claim 1, wherein obtaining the predicted number of uses of each card reading module based on the predicted feature data of each card reading module and the number of uses prediction model of each card reading module comprises:
obtaining the set number of times of use of the card reading module according to the predicted characteristic data of the card reading module and the set number of predicted sub-models;
calculating the predicted use times of the card reading module according to the set number of use times of the card reading module and the weight corresponding to each use time; the weight corresponding to each use frequency of the card reading module is obtained in advance.
3. The method of claim 1 or 2, wherein obtaining the usage prediction model for each card reading module based on the usage training data and the historical usage training for each card reading module comprises:
acquiring the training data of the use times of the card reading module and the historical use times;
acquiring the use frequency characteristic data of the card reading module according to the use frequency training data of the card reading module;
Dividing the usage frequency characteristic data of the card reading module into a training set and a verification set;
training to obtain a preset number of to-be-determined predictive sub-models according to the training set, the historical use times corresponding to the training set and the preset number of preset models; wherein the preset number is greater than or equal to the set number;
verifying the preset number of pending predictor models according to the verification set and the historical use times corresponding to the verification set to obtain an R square value of each pending predictor model;
and selecting the set number of the predictive sub-models from the preset number of the predictive sub-models according to the R square value of each predictive sub-model.
4. The method of claim 3, wherein the obtaining the usage count feature data of the card reading module according to the usage count training data of the card reading module comprises:
performing feature construction on the training data of the using times of the card reading module to obtain multiple types of training feature data;
and selecting the using times characteristic data of the card reading module from the multiple training characteristic data according to a characteristic selection algorithm.
5. A fault early warning apparatus for a card reader device, comprising:
The prediction data acquisition unit is used for acquiring the prediction data of each card reading module of the card reader device; the prediction data of each card reading module is obtained by preprocessing historical record data of a past first preset time period of each card reading module of the card reader device;
the first feature processing unit is used for carrying out feature processing on the prediction data of each card reading module to obtain the prediction feature data of each card reading module; the method comprises the steps that the feature processing of the prediction data of each card reading module comprises feature construction, wherein the feature construction is used for constructing different types of features from the prediction data of each card reading module;
the prediction unit is used for obtaining the predicted use times of each card reading module in a future period of time according to the prediction characteristic data of each card reading module and the use times prediction model of each card reading module; the use number prediction model of each card reading module comprises a set number of predictor models, wherein the set number of predictor models are obtained according to use number training data and historical use number training of each card reading module;
the obtaining unit is used for obtaining the predicted accumulated use times of each card reading module according to the predicted use times and the current accumulated use times in a future period of time of each card reading module; the predicted accumulated use times of the card reading module are equal to the sum of the predicted use times and the current accumulated use times in a future period of time of the card reading module;
And the judging unit is used for sending out fault early warning information after judging that the predicted accumulated use times of the card reading module are larger than the corresponding use threshold.
6. The apparatus of claim 5, wherein the prediction unit comprises:
the prediction subunit is used for obtaining the set number of times of use of the card reading module according to the prediction characteristic data of the card reading module and the set number of prediction submodels;
the calculating subunit is used for calculating the predicted use times of the card reading module according to the set number of use times of the card reading module and the weight corresponding to each use time; the weight corresponding to each use frequency of the card reading module is obtained in advance.
7. The apparatus according to claim 5 or 6, further comprising:
the training data acquisition unit is used for acquiring the training data of the use times of the card reading module and the historical use times;
the second feature processing unit is used for obtaining the use frequency feature data of the card reading module according to the use frequency training data of the card reading module;
the dividing unit is used for dividing the use frequency characteristic data of the card reading module into a training set and a verification set;
The training unit is used for training to obtain a preset number of to-be-determined predictive sub-models according to the training set, the historical use times corresponding to the training set and the preset number of preset models; wherein the preset number is greater than or equal to the set number;
the verification unit is used for verifying the preset number of the undetermined prediction sub-models according to the verification set and the historical use times corresponding to the verification set to obtain an R square value of each undetermined prediction sub-model;
and the selection unit is used for selecting the set number of the predictive sub-models from the preset number of the predictive sub-models according to the R square value of each predictive sub-model.
8. The apparatus of claim 7, wherein the second feature processing unit comprises:
the construction subunit is used for carrying out characteristic construction on the training data of the using times of the card reading module to obtain multiple types of training characteristic data;
and the selecting subunit is used for selecting the using times characteristic data of the card reading module from the multiple types of training characteristic data according to a characteristic selection algorithm.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of claims 1 to 4 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any one of claims 1 to 4.
CN202010585023.9A 2020-06-24 2020-06-24 Fault early warning method and device for card reader equipment Active CN111783487B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010585023.9A CN111783487B (en) 2020-06-24 2020-06-24 Fault early warning method and device for card reader equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010585023.9A CN111783487B (en) 2020-06-24 2020-06-24 Fault early warning method and device for card reader equipment

Publications (2)

Publication Number Publication Date
CN111783487A CN111783487A (en) 2020-10-16
CN111783487B true CN111783487B (en) 2023-08-25

Family

ID=72759749

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010585023.9A Active CN111783487B (en) 2020-06-24 2020-06-24 Fault early warning method and device for card reader equipment

Country Status (1)

Country Link
CN (1) CN111783487B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112464563A (en) * 2020-11-27 2021-03-09 河北建设投资集团有限责任公司 Data mining method for steam turbine fault diagnosis

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0589315A (en) * 1991-09-30 1993-04-09 Omron Corp Control equipment
CN104361404A (en) * 2014-10-28 2015-02-18 哈尔滨工业大学 Airborne wireless data communication relay unit residual-life prediction method and device
CN106740801A (en) * 2016-12-13 2017-05-31 中车株洲电力机车有限公司 A kind of monitoring method of track train braking control system
CN207282429U (en) * 2017-07-21 2018-04-27 安徽信息工程学院 Breaker service life precaution device
CN110623679A (en) * 2019-09-20 2019-12-31 四川九八村信息科技有限公司 System and method for predicting equipment failure of plasma apheresis machine
CN111049571A (en) * 2019-12-29 2020-04-21 苏州浪潮智能科技有限公司 Optical module fault prediction method and device and computer readable storage medium
CN111078446A (en) * 2019-11-22 2020-04-28 北京达佳互联信息技术有限公司 Fault information acquisition method and device, electronic equipment and storage medium

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0589315A (en) * 1991-09-30 1993-04-09 Omron Corp Control equipment
CN104361404A (en) * 2014-10-28 2015-02-18 哈尔滨工业大学 Airborne wireless data communication relay unit residual-life prediction method and device
CN106740801A (en) * 2016-12-13 2017-05-31 中车株洲电力机车有限公司 A kind of monitoring method of track train braking control system
CN207282429U (en) * 2017-07-21 2018-04-27 安徽信息工程学院 Breaker service life precaution device
CN110623679A (en) * 2019-09-20 2019-12-31 四川九八村信息科技有限公司 System and method for predicting equipment failure of plasma apheresis machine
CN111078446A (en) * 2019-11-22 2020-04-28 北京达佳互联信息技术有限公司 Fault information acquisition method and device, electronic equipment and storage medium
CN111049571A (en) * 2019-12-29 2020-04-21 苏州浪潮智能科技有限公司 Optical module fault prediction method and device and computer readable storage medium

Also Published As

Publication number Publication date
CN111783487A (en) 2020-10-16

Similar Documents

Publication Publication Date Title
CN109495318B (en) Mobile communication network flow prediction method, device and readable storage medium
US11107014B2 (en) Method and system for model fitting to hierarchical time series cluster
CN108959934A (en) Safety risk estimating method, device, computer equipment and storage medium
CN102567391A (en) Method and device for building classification forecasting mixed model
CN103778474A (en) Resource load capacity prediction method, analysis prediction system and service operation monitoring system
CN109583629A (en) Improved similar historical day short-term load forecasting method and device based on deviation self-correcting
US20170124495A1 (en) Method and system for mitigating risk in a supply chain
CN111177485B (en) Parameter rule matching based equipment fault prediction method, equipment and medium
CN111783487B (en) Fault early warning method and device for card reader equipment
US11265688B2 (en) Systems and methods for anomaly detection and survival analysis for physical assets
Dhillon et al. Peak electricity load forecasting using online support vector regression
CN110633863A (en) Bank note distribution prediction method and device based on GBDT algorithm
CN111105050B (en) Fan maintenance plan generation method, device, equipment and storage medium
US11500340B2 (en) Performance evaluation based on resource dynamics
CN116645132A (en) Multi-factor variable-based time sequence prediction method and device, electronic equipment and medium
CN111783486B (en) Maintenance early warning method and device for card reader equipment
CN109978299B (en) Data analysis method and device for offshore wind power business and storage medium
CN115577996A (en) Risk assessment method, system, equipment and medium for power grid power failure plan
Mohammadi et al. Machine learning assisted stochastic unit commitment: A feasibility study
CN114841664A (en) Method and device for determining multitasking sequence
US20210027234A1 (en) Systems and methods for analyzing user projects
CN112637793A (en) Scene charging method, system, electronic equipment and storage medium based on 5G
CN111738507A (en) Bank clearing position fund payment amount prediction method, device, equipment and medium
CN111950753A (en) Scenic spot passenger flow prediction method and device
CN114997879B (en) Payment routing method, device, equipment and storage medium

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