WO2020248844A1 - 测试对象的寿命预估方法、装置、设备及介质 - Google Patents

测试对象的寿命预估方法、装置、设备及介质 Download PDF

Info

Publication number
WO2020248844A1
WO2020248844A1 PCT/CN2020/093362 CN2020093362W WO2020248844A1 WO 2020248844 A1 WO2020248844 A1 WO 2020248844A1 CN 2020093362 W CN2020093362 W CN 2020093362W WO 2020248844 A1 WO2020248844 A1 WO 2020248844A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
test object
life
type
test
Prior art date
Application number
PCT/CN2020/093362
Other languages
English (en)
French (fr)
Inventor
谢静文
金晓辉
阮晓雯
徐亮
Original Assignee
平安科技(深圳)有限公司
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 平安科技(深圳)有限公司 filed Critical 平安科技(深圳)有限公司
Publication of WO2020248844A1 publication Critical patent/WO2020248844A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2455Query execution
    • G06F16/24564Applying rules; Deductive queries
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2458Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/04Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"

Definitions

  • This application relates to the field of data analysis in the field of big data technology, and specifically to a method, device, equipment, and medium for predicting the life of a test object.
  • Life expectancy is a way to predict the life of test objects (for example, individuals or products such as equipment).
  • the inventor realizes that in the process of traditional life expectancy, the evaluation of test subjects often uses questionnaire surveys to subjectively judge whether the test subjects are disabled, but questionnaire surveys have problems such as uneven sampling, small sample size, and large subjective judgments. , Will cause a large deviation in the estimated results.
  • the sample of test subjects in the questionnaire survey is not comprehensive, and the relevant information of the test subjects filled in the questionnaire survey is also easy to fill in errors, which will adversely affect the accuracy and calculation efficiency of life expectancy; and pass
  • the questionnaire survey method to estimate the life span of the test subject requires a lot of manpower and material resources, resulting in waste of resources. Therefore, it is an urgent need for those skilled in the art to find a technical solution that can solve the above-mentioned problems.
  • the embodiments of the present application provide a method, device, equipment, and medium for estimating the life of a test object, which saves costs, improves the fine-grained life estimation, and improves data matching efficiency and life estimation in the life estimation process. Precision and life expectancy efficiency.
  • a method for estimating the life of test objects including:
  • Receive a data analysis instruction including the initial parameters of each type of the test object, input each of the specified item loss coefficients and each of the initial parameters into a preset life loss analysis model, and receive the life loss analysis model Output the cumulative lost life of each type of test object;
  • the initial life expectancy and the accumulated lost life of each type of the test object are input into a preset adjustment model, and the adjusted life expectancy of each type of the test object output by the adjustment model is received.
  • a device for predicting the life of a test object including:
  • the data retrieval module is used to receive data retrieval instructions containing each test group, and obtain the object data table associated with each test group from the database; one test group corresponds to one type of test object;
  • the data matching module is configured to receive a data matching instruction containing a designated item identifier, and count the number of valid data matching the designated item identifier from each of the object data tables;
  • a data determining module configured to determine the specified item loss coefficient of each type of the test object according to the number of valid data matching the specified item identifier
  • the data analysis module is used to receive data analysis instructions including the initial parameters of each type of test object, input the loss coefficients of each specified item and the initial parameters into a preset life loss analysis model, and receive The cumulative lost life of each type of the test object output by the life loss analysis model;
  • the data adjustment module is used to input the initial expected life and the accumulated lost life of each type of the test object into a preset adjustment model, and receive the adjustment expectation of each type of the test object output by the adjustment model life.
  • a computer device includes a memory, a processor, and computer-readable instructions stored in the memory and capable of running on the processor, and the processor implements the following steps when the processor executes the computer-readable instructions:
  • Receive a data analysis instruction including the initial parameters of each type of the test object, input each of the specified item loss coefficients and each of the initial parameters into a preset life loss analysis model, and receive the life loss analysis model Output the cumulative lost life of each type of test object;
  • the initial life expectancy and the accumulated lost life of each type of the test object are input into a preset adjustment model, and the adjusted life expectancy of each type of the test object output by the adjustment model is received.
  • One or more readable storage media storing computer readable instructions, when the computer readable instructions are executed by one or more processors, the one or more processors execute the following steps:
  • Receive a data analysis instruction including the initial parameters of each type of the test object, input each of the specified item loss coefficients and each of the initial parameters into a preset life loss analysis model, and receive the life loss analysis model Output the cumulative lost life of each type of test object;
  • the initial life expectancy and the accumulated lost life of each type of the test object are input into a preset adjustment model, and the adjusted life expectancy of each type of the test object output by the adjustment model is received.
  • This application saves costs, and uses more abundant test group data, improves the fine-grained life expectancy, and improves data matching efficiency, life expectancy accuracy, and life expectancy efficiency in the life estimation process; at the same time; Objectively reflects the life of the test object.
  • FIG. 1 is a schematic diagram of an application environment of a method for predicting the life of a test object in an embodiment of the present application
  • FIG. 2 is a flowchart of a method for estimating the life of a test object in an embodiment of the present application
  • step S10 of the method for estimating the life of a test object in an embodiment of the present application
  • step S20 is a flowchart of step S20 of the method for estimating the life of a test object in an embodiment of the present application
  • step S40 of the method for estimating the life of a test object in an embodiment of the present application
  • step S40 of the method for estimating the life of a test object in another embodiment of the present application.
  • FIG. 7 is a schematic block diagram of a device for predicting the life of a test object in an embodiment of the present application.
  • Fig. 8 is a schematic diagram of a computer device in an embodiment of the present application.
  • This application relates to the field of big data technology.
  • the method for estimating the life of a test object provided in this application can be applied to the application environment as shown in FIG.
  • the client includes, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, cameras, and portable wearable devices.
  • the server can be implemented as an independent server or a server cluster composed of multiple servers.
  • a method for estimating the lifespan of a test object is provided. Taking the method applied to the server in FIG. 1 as an example for description, the method includes the following steps:
  • S10 Receive a data retrieval instruction including each test group, and obtain an object data table associated with each test group from a database; one test group corresponds to one type of test object.
  • the data retrieval instruction refers to a plurality of consecutive test groups input by the user on the client, and the user clicks a preset button bound to the data retrieval instruction and then sent to the server;
  • the test group is Divided according to a preset period (for example, 1 year);
  • the test object may be a single product that needs to predict the service life, the test object may be an individual who needs to predict a healthy life;
  • the object data is Refers to the direct data affecting the life of the test object.
  • each test group can have a 0-year service life, a 1-year service life, a 2-year service life, a 3-year service life, etc.
  • the object data may be used to record the usage status of the test object
  • the maintenance data that is, the data for regular inspection, daily maintenance or fault maintenance of the product
  • the maintenance data includes but is not limited to the product name, product maintenance order number, maintenance date, maintenance items, etc.
  • each test group may be 0 years old, 1 year old, 2 years old, 3 years old, etc.
  • the object data may be case data used to record the health status of the individual (also called Personal medical record data)
  • the case data includes, but is not limited to, the name of the diagnosed person, the document number of the medical record and medical record file, the date of diagnosis, the name of the disease, and the disease code.
  • the server when it receives the data retrieval instruction including each test group, it obtains the test object table associated with each test group from the database, and the test object table contains the object data of each test object in the same test group. .
  • the server can be applied to a life estimation system.
  • S20 Receive a data matching instruction containing a designated item identifier, and count the number of valid data matching the designated item identifier from each of the object data tables.
  • the data matching instruction refers to a user uploading a locally stored designated item file or entering a designated item identifier on the client, and clicking a preset button bound to the data matching instruction, and then sending it to the server;
  • the designated item file contains a preset number of designated items, designated item identifiers corresponding to the designated items, adjustment weights corresponding to the designated items, etc.; the designated item identifiers include, but are not limited to, the project name, project code, or project key of the designated item word.
  • the preset application program reads all the designated item identifiers in the designated item file, and then sends all the designated item identifiers to the server.
  • the server Receive the data matching instruction containing the specified item identification, use the dual index matching method to synchronize the query from each object data table for the number of object data matching the specified item identification, and according to the number of object data matching the specified item identification Determine the number of valid data in each object data table.
  • the dual-index matching method refers to matching the names or codes contained in each object data in the object data table by using a regular matching method.
  • S30 Determine the specified item loss coefficient of each type of the test object according to the number of valid data matching the specified item identifier.
  • the specified item loss coefficient is used to calculate the lost life of each type of test object.
  • the number of valid data matching the specified item identifier is used as an input parameter, input into the coefficient model associated with the specified item (that is, one index item is associated with one coefficient model), and the specified item loss coefficient output by the coefficient model is received .
  • the coefficient model is:
  • Fx(i) is the loss coefficient of the specified item
  • Ex(i) is the number of valid data matching the specified item identifier
  • w(i) is the adjustment weight corresponding to the specified item, in the coefficient model Is a constant
  • Mx is the amount of data in the object data table. Understandably, according to the above coefficient model The probability of the specified item can be obtained.
  • the data analysis instruction means that the user uploads a locally stored lifespan file or inputs the initial parameters of each type of test object on the client, and then clicks the preset button bound to the data analysis instruction to be sent to Server;
  • the life file contains the initial number of objects of each type of test object, the number of marked objects, the marked rate, average life, total life, initial predicted life and other initial parameters;
  • the marked rate refers to the entered The probability that a type of test object of an exact test group will be marked within the test duration of the test group;
  • the initial number of objects refers to the number of unmarked test objects of a type of test object that have just entered a certain test group;
  • the number of marked objects refers to the number of marked test objects of a type of test object in a certain test group during the test period;
  • the average life span refers to the number of test objects that enter an exact test group before entering another exact test group.
  • the length of time that may not be marked; the total life span refers to a type of test object that enters the exact test group, and the total length of time that the type of test object may not be marked in the future is determined according to a certain marked rate; the initial life expectancy Refers to a type of test object entering a certain test group, according to a certain marked rate to determine the average length of time that the type of test object may not be marked in the future.
  • the life file is a data table stored in a table format model.
  • the data table may refer to obtaining an initial number of objects (for example, 100,000), and determining each type of test object according to the number of marked objects.
  • the marked rate of the similar test objects further constitutes a tabular model until the difference between the initial number of objects and the marked number is less than or equal to a preset threshold (for example, 0). Understandable,
  • the life file contains the marked rate of the product in each life cycle from production to obsolescence (that is, the product is marked as obsolete Probability), the marked rate in each life cycle can be used as an important reference basis for users to decide whether to purchase a product, and it can also reflect the elimination rule of a product at a certain test point.
  • the data table contains the marked rate of the individual from birth to death (that is, the probability that the individual is marked as dead) ), the marked rate in each age group can be used as an important basis for users to apply for insurance, and can also reflect the individual health status and the law of survival and death at a certain test point.
  • the preset application reads the initial parameters of each type of test object in the lifespan file, and then sends the initial parameters of each type of test object to In the server, at this time, the server receives a data analysis instruction containing the initial parameters of each type of test object, and uses the initial parameters of each type of test object and the specified item loss coefficient group in the above step S30 as input parameters, and inputs them to the prediction Set the life loss analysis model, and receive the cumulative lost life output of each type of test object output by the life loss analysis model, and then store each type of test object in association with the corresponding cumulative lost life in the database.
  • the life loss analysis model uses the data in the coefficient array and the parameter array to process iteratively, accumulate, etc., to obtain an output array containing the cumulative lost life of each type of test object to improve data processing efficiency and avoid data errors .
  • S50 Input the initial life expectancy and the accumulated lost life of each type of the test object into a preset adjustment model, and receive the adjusted life expectancy of each type of the test object output by the adjustment model.
  • the adjustment model is:
  • Adj ⁇ ex is the adjusted life expectancy
  • ex is the initial life expectancy
  • Total ⁇ Lx is the cumulative lost life.
  • the initial life expectancy included in the initial parameters and the cumulative lost life in step S40 are used as input parameters, input into the adjustment model, and the adjusted life expectancy output from the adjustment model is received, and then each type of test object The corresponding adjusted life expectancy is associated and stored in the database. At this time, the cumulative lost life and adjusted life expectancy of each type of test object are also associated and stored.
  • the adjusted life expectancy of each type of the test object when the adjusted life expectancy of each type of the test object is obtained, it is detected whether the adjusted life expectancy of the test object reaches an alarm condition, and whether the adjusted life expectancy of the test object reaches When an alarm condition occurs, an early warning message is sent to the client.
  • this embodiment obtains the object data tables of each test group from the database according to the data retrieval instruction, and counts the number of valid data matching the specified item identifier from each object data table according to the data matching instruction, thereby determining each category The loss coefficient of the specified item of the test object.
  • the initial parameters and the loss coefficient of the specified item are input into the life loss analysis model according to the data analysis instruction, and the cumulative lost life output by the life loss analysis model is received, and then according to the The initial life expectancy and cumulative lost life are determined and adjusted.
  • This embodiment forms a system reusable life estimation method, which saves costs, and uses more abundant test group data, which improves the fine-grained life estimation. And in the life estimation process, the data matching efficiency, the life estimation accuracy and the life estimation efficiency are improved, and it is conducive to the follow-up life monitoring; at the same time, it objectively reflects the life of the test object.
  • the method for estimating the life of the test object further includes the following steps:
  • a login request containing a user ID is received from the client, and after the identity verification of the logged-in user is passed, the query permission level of the logged-in user is obtained from a preset user permission table according to the user ID.
  • a query instruction containing the query authority level and query conditions is received, query data that matches the query authority level and meets the query conditions is obtained, and the query data is exported in a preset document form and sent to the Client; the query data includes the adjusted life expectancy of each type of test object.
  • the user authority table includes a user identifier and a query authority level corresponding to the user identifier; the user identifier includes but is not limited to a user name and a user number; the query authority level is set according to requirements, for example: 1 ⁇ Level 5.
  • the query conditions include the test point where the query object is located, the test group where the query object is located, and the type of the query object.
  • each test point and the object data table have been associated and stored in the database.
  • the login instruction is received from the client, and the login instruction contains the user ID of the logged-in user.
  • the user table (containing one or more valid user IDs and pre-stored in the database) is used to query whether there is a user ID. If there is a valid user ID that matches the user ID in the user table, it is determined that the authentication of the logged-in user is passed; and when there is no valid user ID that matches the user ID in the user table, the ID of the logged-in user is determined Identity verification failed.
  • the server receives the query instruction including the query condition and the query authority level, searches the database for query data that is associated with the query authority level and meets the query condition, and combines the query data, query time, and query conditions.
  • the query authority level, etc. are added to the preset data table template to generate the query data table, and the query data table is exported and sent to the client, so that the user can view, edit, and analyze the query data.
  • this embodiment receives and verifies the login request sent by the logged-in user, obtains the query authority level of the logged-in user from the user authority table, and then queries and obtains the query authority level that matches the query authority level and meets the query conditions according to the query instruction.
  • Query data to avoid cross-authorization operations by logged-in users, ensuring data security and reliability.
  • the step S10 specifically includes the following steps:
  • test object set of each test group from a preset data center, one test object set includes at least one test object of the same type, and each test object includes at least one object document.
  • the target document may refer to a paper scan that records the target data and is stored in a data center, and the target document may be stored in a label image file format, a picture format, or a portable document format.
  • a preset data center for example, a product maintenance center set for the product for the test subject, an electronic medical record center set for the individual for the test subject, etc.
  • the information is obtained from the data center for the test object collections of each test group.
  • the number of test objects and the number of test documents contained in each test object collection are the same;
  • the authorization information includes the authorization number of the data center;
  • the retrieval information Including the number of selected target documents, the order of selecting the target documents (for example, according to the storage time of the target documents), etc., for example, randomly selecting 10,000 target documents associated with a certain test group.
  • S102 Allocate all the object documents in the same test object set to the same data extraction thread. That is, one data extraction thread is used to identify all the object documents in a test object collection, and then extract the object data contained in the object documents. Understandably, multiple data extraction threads can identify and extract data synchronously, which is convenient for data Storage, and improve the efficiency of data processing.
  • S103 Invoke an optical character recognition model associated with the data extraction thread, identify all the target documents allocated to the data extraction thread, and obtain information contained in each target document in the data extraction thread.
  • Object data That is, a data extraction thread is associated with an optical character recognition model, and the optical character recognition model is used to identify all the object documents allocated to the data extraction thread, and then the object data contained in each object document is extracted. Including but not limited to the basic information of the detected object in the target document, document number, document date, document source information, and detected item.
  • the object document is recognized by the optical character recognition model, which has faster timeliness and lower error rate; and the recognition process does not require manual operation, and does not cause a time interval for manual operation to be stopped, and at the same time Multiple target documents are recognized, and the recognition efficiency is higher.
  • the optical recognition character recognition model is a learning model generated based on the training of the document to be recognized.
  • the step S103 includes the following steps: acquiring the document to be recognized, and training to generate optical character recognition based on the document to be recognized model.
  • the document to be recognized is a historical object document stored in the data center.
  • training the optical character recognition model 2000 objects of the same type can be used as the document to be recognized. After the document is learned, it is necessary to correct the results according to the learning content, and after repeated learning, an optical character recognition model that can target the document is generated.
  • S104 Generate an object data table according to the object data contained in each object document in the data extraction thread, and store the object data table in association with the corresponding test group in the database. That is, all the object data extracted by the same data extraction thread is added to a data table model to generate the object data table, and the object data table and the corresponding test group are associated and stored in the database. Understandably, the object data table can be updated according to user requirements.
  • this embodiment obtains multiple test object sets, and uses multiple data extraction threads to synchronize all the object documents in each test object set to facilitate data storage, improve recognition efficiency, and extract multiple data
  • the threads recognize each other to avoid data interference; and the optical character recognition model recognizes the target document, which has faster timeliness and lower error rate.
  • the entire recognition process does not require time intervals, which further improves the recognition efficiency; at the same time, the test group has rich data. Improved the fine-grained life expectancy.
  • step S20 specifically includes the following steps:
  • S201 Receive a data matching instruction, where the data matching instruction includes the item name and item code of each designated item.
  • the received data matching instruction may also include item keywords of designated items, etc., through the two regular matching methods of the name matching expression in step S202 and the code matching expression in step S203, Finally, in the object data table, the successfully matched object data is queried and marked, and the successfully matched object data is recorded as valid object data; understandably, as long as one of the two matching methods meets the requirements, the matching is successful.
  • test subject is an individual, and the acquired designated item identifier (that is, the disease name or keyword of the designated disease) is "stroke”, use "stroke” to search in the national disease classification standard code, and first lock the disease
  • the first three digits of the standard code are all diseases with I60-I69, some examples are shown in Table 1, and then some disease names in Table 1 are summarized, such as all including "infarction”, “infarction” or “hemiplegia”. Words to form a text matching summary of the disease name.
  • stroke patients are divided into first stroke and long-term stroke patients.
  • the first three digits in the disease standard code are I69 corresponding to all stroke sequelae (as shown in Table 2 below), and the name matching expression has also been modified accordingly ;
  • this embodiment sets the matching standard of the object data through the dual-index matching method, precisely determines the test object that needs to be marked, and achieves the purpose of accurate data matching and efficient data matching.
  • step S30 specifically includes the following steps:
  • each designated item corresponds to a designated item identifier and disease weight.
  • the weight can be determined and adjusted from the designated item file according to the designated item identifier.
  • the data amount of each test data table is obtained. That is, the number of all object data in each object data table.
  • the coefficient model corresponding to each designated item identifier can be determined according to the adjustment weight and the total number of objects in the above steps (refer to step S30), and the coefficient model corresponding to each designated item identifier can be The number of valid data matching the designated item identifier is input into the coefficient model corresponding to each designated item identifier, and the loss coefficient of each designated item of each type of test object can be obtained.
  • the present embodiment adjusts the initial life expectancy based on the specified item loss coefficient, which is beneficial to objectively reflect the life of the test object at each test point.
  • step S40 specifically includes the following steps:
  • S401 Receive a data analysis instruction, where the data analysis instruction includes the initial expected life span, the initial number of objects, and the total life span of each type of the test object.
  • the data analysis instruction may also include the number of marked objects, the marked rate, average lifespan, etc., wherein the relationship between the initial number of objects, total lifespan, and initial life expectancy of each type of test object is as follows:
  • Tx is the total lifetime
  • Ix is the initial number of objects.
  • Mr is the marked rate
  • Mx is the number of marked objects.
  • S402 Iterate the initial object quantity, the total life span, and the loss coefficient of each specified item of each type of the test object through the life loss analysis model to obtain the specified item of each type of the test object Lost life.
  • the steps for obtaining the specified life loss of each type of test object are as follows:
  • the specified item lost life of each type of test object is obtained.
  • Lx(i) is the specified item lost life of each type of test object
  • SUM(Fx(i)*Tx) is the specified item loss life of each type of test object Total life lost.
  • S403 Accumulate the specified item lost life of each type of the test object through the life loss analysis model to obtain the cumulative lost life of each type of the test object.
  • the cumulative sub-model in the life loss analysis model is used to accumulate the lost life of all index items (a designated item corresponds to a designated item lost life) to obtain the cumulative lost life.
  • the cumulative sub-model is:
  • n is the number of specified items. Understandably, when there is one designated item, the lost life of the designated item will be regarded as the cumulative lost life.
  • the present embodiment obtains the cumulative lost life through the life loss analysis model, and automatically calculates it, which improves the data processing efficiency and at the same time improves the accuracy of the lost life.
  • S404 Determine the marked rate of each type of test object according to the acquired number of test objects and the number of marked objects of each type of the test object.
  • the relationship between the number of test objects, the number of marked objects, and the marked rate is known (refer to step S401).
  • the forward iteration model is:
  • I j-1 I j-2 -I j-2 *Mr j-2
  • I j-1 is the initial number of test objects of a type corresponding to the next test group
  • I j-2 is the initial number of test objects of a type corresponding to the previous test group
  • the steps for obtaining the total life span of each type of test object by using the reverse iteration model are as follows:
  • T w-1 ⁇ I w-1 *Q; where T w-1 is the total life of a type of test object corresponding to the highest test group. Life, I w-1 is the initial number of test objects of a type corresponding to the highest test group, and Q represents the cycle time of the test group, for example, 1 year.
  • the relationship between the number of initial objects, total lifespan, and initial life expectancy of each type of test object is known (refer to step S401).
  • it can be based on the initial object of each type of test object.
  • the number, the total life span and the initial expected life span generate a life file and store the life file in the database.
  • the server only needs to obtain the life file from the database.
  • the initial parameters acquired in this embodiment have high richness and high accuracy, which is beneficial to improve the accuracy of life estimation.
  • a device for predicting the life of a test object is provided, and the device for predicting the life of the test object corresponds to the method for predicting the life of the test object in the foregoing embodiment.
  • the life estimation device of the test object includes the following modules, and each functional module is described in detail as follows:
  • the data retrieval module 110 is configured to receive data retrieval instructions including each test group, and obtain an object data table associated with each test group from a database; one test group corresponds to one type of test object.
  • the data matching module 120 is configured to receive a data matching instruction including a designated item identifier, and count the number of valid data matching the designated item identifier from each of the object data tables.
  • the data determining module 130 is configured to determine the specified item loss coefficient of each type of the test object according to the number of valid data matching the specified item identifier.
  • the data analysis module 140 is configured to receive a data analysis instruction including the initial parameters of each type of the test object, and input the loss coefficients of each specified item and the initial parameters into a preset life loss analysis model, and Receive the cumulative lost life of each type of the test object output by the life loss analysis model.
  • the data adjustment module 150 is configured to input the initial expected life and the accumulated lost life of each type of the test object into a preset adjustment model, and receive the adjustment of each type of the test object output by the adjustment model Life expectancy.
  • the device for predicting the life of the test object further includes the following modules, and each functional module is described in detail as follows:
  • the login module is configured to receive a login request containing a user ID sent from the client, and obtain the query permission level of the logged-in user from a preset user permission table according to the user ID after the login user's identity verification is passed.
  • the query display model is used to receive a query instruction including the query authority level and query conditions, obtain query data that matches the query authority level and meets the query conditions, and export the query data in the form of a preset document. Sent to the client; the query data includes the adjusted life expectancy of each type of the test object.
  • the data retrieval module 110 includes the following sub-modules, and each functional sub-module is described in detail as follows:
  • the calling sub-module is used to obtain the test object set of each test group from a preset data center.
  • One test object set includes at least one test object of the same type, and each test object includes at least one Object document.
  • the allocation sub-module is used to allocate all the object documents in the same test object collection to the same data extraction thread.
  • the calling sub-module is used to call the optical character recognition model associated with the data extraction thread, identify all the object documents allocated to the data extraction thread, and obtain each object in the data extraction thread Object data contained in the document.
  • the first storage submodule is configured to generate an object data table based on the object data contained in each object document in the data extraction thread, and store the object data table in association with the corresponding test group in the In the database.
  • the data matching module 120 includes the following sub-modules, and each functional sub-module is described in detail as follows:
  • the matching sub-module is used to receive a data matching instruction, the data matching instruction including the item name and item code of each designated item.
  • the name sub-module is used to obtain the corresponding name matching expression according to the project name of each specified item.
  • the encoding sub-module is used to obtain the corresponding encoding matching expression according to the item encoding of each specified item.
  • the marking sub-module is used to obtain and mark the object data satisfying the name matching expression and/or the code matching expression from each of the object data tables, and collect statistics on the marked object data to obtain data from each
  • the item name and/or each item code refers to the number of matched valid data.
  • the data determining module 130 includes the following sub-modules, and each functional sub-module is described in detail as follows:
  • the first obtaining submodule is used to obtain the adjustment weight associated with the specified item identifier.
  • the second acquisition sub-module is used to acquire the data volume of each test data table.
  • the third acquisition sub-module is used to acquire the loss of the designated item of each type of the test object according to the data volume of each test data table, the number of valid data matching each designated item identifier, and the associated adjustment weight coefficient.
  • the data analysis module 140 includes the following sub-modules, and each functional sub-module is described in detail as follows:
  • the analysis sub-module is used to receive data analysis instructions, the data analysis instructions including the initial expected life, the number of initial objects and the total life of each type of the test object.
  • the iterative sub-module is used to iterate the initial object quantity, the total life span, and the loss coefficient of each specified item of each type of the test object through the life loss analysis model to obtain each type of the test The specified item of the object loses its life.
  • the accumulation sub-module is used to accumulate the specified item lost life of each type of test object through the life loss analysis model to obtain the accumulated lost life of each type of test object.
  • the data analysis module 140 further includes the following sub-modules, and each functional sub-module is described in detail as follows:
  • the fourth acquisition sub-module is used to determine the marked rate of each type of test object according to the number of test objects and the number of marked objects of each type of test object obtained;
  • the forward iteration sub-module is used to input the acquired initial number of test objects of one type of the test object corresponding to the initial test group and the marked rate of each type of the test object into the preset forward iteration model , And receive the initial object quantity of each type of the test object output by the forward iterative model;
  • the reverse iteration sub-module is used to input the initial number of objects of each type of the test object into a preset reverse iteration model, and receive each type of the test object output by the reverse iteration model Total life span;
  • the second storage sub-module is used to obtain the initial expected life of each type of test object according to the initial object number and the total life of each type of the test object, and to compare the initial object number, The total life and the initial expected life are stored in the database in association.
  • the various modules in the device for predicting the life of the test object can be implemented in whole or in part by software, hardware, and combinations thereof.
  • the foregoing modules may be embedded in the form of hardware or independent of the processor in the computer device, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the foregoing modules.
  • a computer device is provided.
  • the computer device may be a server, and its internal structure diagram may be as shown in FIG. 8.
  • the computer equipment includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide calculation and control capabilities.
  • the memory of the computer device includes a readable storage medium and an internal memory.
  • the readable storage medium stores an operating system, computer readable instructions, and a database.
  • the internal memory provides an environment for the operation of the operating system and computer readable instructions in the readable storage medium. When the computer-readable instructions are executed by the processor, a method for estimating the life of the test object is realized.
  • the readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.
  • a computer device including a memory, a processor, and computer-readable instructions stored in the memory and capable of running on the processor, and the processor implements the following steps when the processor executes the computer-readable instructions:
  • Receive a data analysis instruction including the initial parameters of each type of the test object, input each of the specified item loss coefficients and each of the initial parameters into a preset life loss analysis model, and receive the life loss analysis model Output the cumulative lost life of each type of test object;
  • the initial life expectancy and the accumulated lost life of each type of the test object are input into a preset adjustment model, and the adjusted life expectancy of each type of the test object output by the adjustment model is received.
  • one or more readable storage media storing computer readable instructions are provided.
  • the readable storage media provided in this embodiment include non-volatile readable storage media and volatile readable storage Medium; the readable storage medium stores computer readable instructions, and when the computer readable instructions are executed by one or more processors, the one or more processors implement the following steps:
  • Receive a data analysis instruction including the initial parameters of each type of the test object, input each of the specified item loss coefficients and each of the initial parameters into a preset life loss analysis model, and receive the life loss analysis model Output the cumulative lost life of each type of test object;
  • the initial life expectancy and the accumulated lost life of each type of the test object are input into a preset adjustment model, and the adjusted life expectancy of each type of the test object output by the adjustment model is received.
  • Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Road DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Computational Linguistics (AREA)
  • Economics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Probability & Statistics with Applications (AREA)
  • Development Economics (AREA)
  • Fuzzy Systems (AREA)
  • Game Theory and Decision Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种测试对象的寿命预估方法、装置、设备及介质。所述方法包括:接收包含各测试组的数据调取指令,自数据库获取与各测试组关联的对象数据表,一个所述测试组对应一类测试对象(S10);接收包含指定项标识的数据匹配指令,自各对象数据表统计与指定项标识匹配的有效数据个数(S20);根据与指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数(S30);接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命(S40);将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命(S50)。本方法节约了成本,提高了寿命预估的细粒度,且在寿命预估过程中提高了数据匹配效率、寿命预估精准度和寿命预估效率。

Description

测试对象的寿命预估方法、装置、设备及介质
本申请要求于2019年6月14日提交中国专利局、申请号为201910516569.6,名称为“测试对象的寿命预估方法、装置、设备及介质”的中国申请专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及大数据技术领域的数据分析领域,具体涉及一种测试对象的寿命预估方法、装置、设备及介质。
背景技术
寿命预估是对测试对象(比如说人员个体或者设备等产品)的寿命进行预测的一种方式。发明人意识到,在传统预测寿命过程中,对测试对象进行评估往往采用问卷调查的方式来主观评判测试对象是否失能,但问卷调查存在取样不均衡,样本量小和主观判断影响大等问题,会导致预估结果产生较大偏差。同时,问卷调查的方式的测试对象的样本并不全面,且问卷调查中的填写的测试对象的相关信息也容易被填写错误,进而对寿命预估的精准度和计算效率产生不良影响;且通过问卷调查对测试对象的寿命进行预估的方式需要耗费大量的人力物力,造成了资源浪费。因此,需寻找一种能解决上述问题的技术方案成为本领域技术人员的迫切需求。
申请内容
本申请实施例提供一种测试对象的寿命预估方法、装置、设备及介质,节约了成本,提高了寿命预估的细粒度,且在寿命预估过程中提高了数据匹配效率、寿命预估精准度和寿命预估效率。
一种测试对象的寿命预估方法,包括:
接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
一种测试对象的寿命预估装置,包括:
数据调取模块,用于接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
数据匹配模块,用于接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
数据确定模块,用于根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
数据分析模块,用于接收包含每一类所述测试对象的初始参数的数据分析指令,将各所 述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
数据调整模块,用于将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
一种计算机设备,包括存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机可读指令,所述处理器执行所述计算机可读指令时实现如下步骤:
接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
一个或多个存储有计算机可读指令的可读存储介质,所述计算机可读指令被一个或多个处理器执行时,使得所述一个或多个处理器执行如下步骤:
接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
本申请节约了成本,且使用的测试组数据更为丰富,提高了寿命预估的细粒度,且在寿命预估过程中提高了数据匹配效率、寿命预估精准度和寿命预估效率;同时客观反映了测试对象的寿命情况。本申请的一个或多个实施例的细节在下面的附图和描述中提出,本申请的其他特征和优点将从说明书、附图以及权利要求变得明显。
附图说明
为了更清楚地说明本申请实施例的技术方案,下面将对本申请实施例的描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本申请一实施例中测试对象的寿命预估方法的应用环境示意图;
图2是本申请一实施例中测试对象的寿命预估方法的流程图;
图3是本申请一实施例中测试对象的寿命预估方法的步骤S10的流程图;
图4是本申请一实施例中测试对象的寿命预估方法的步骤S20的流程图;
图5是本申请一实施例中测试对象的寿命预估方法的步骤S40的流程图;
图6是本申请另一实施例中测试对象的寿命预估方法的步骤S40的流程图;
图7是本申请一实施例中测试对象的寿命预估装置的原理框图;
图8是本申请一实施例中计算机设备的示意图。
具体实施方式
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
本申请涉及大数据技术领域,本申请提供的测试对象的寿命预估方法,可应用在如图1的应用环境中,其中,客户端通过网络与服务器进行通信。其中,客户端包括但不限于为各种个人计算机、笔记本电脑、智能手机、平板电脑、摄像头和便携式可穿戴设备。服务器可以用独立的服务器或者是多个服务器组成的服务器集群来实现。
在一实施例中,如图2所示,提供一种测试对象的寿命预估方法,以该方法应用在图1中的服务器为例进行说明,包括以下步骤:
S10,接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象。
在本实施例中,所述数据调取指令是指用户在客户端输入的连续多个测试组,并点击与数据调取指令绑定的预设按钮之后发送至服务器的;所述测试组是根据预设周期时长(例如,1年)进行划分的;所述测试对象可以是需要预测使用寿命的单个产品,所述测试对象可以为亦可以是需要预测健康寿命的个体;所述对象数据是指影响测试对象寿命的直接数据。
示例性的,若测试对象为产品,则各测试组可以为0年使用周期、1年使用周期、2年使用周期、3年使用周期等,而对象数据可以为用于记载测试对象的使用状况的检修数据(也即对产品进行定期检查、日常维修或故障维修的数据),所述检修数据包括但不限定于产品名称、产品检修单的编号、检修日期、检修项等。
示例性的,若测试对象为个体,则各测试组可以为0岁段、1岁段、2岁段、3岁段等,而对象数据可以为用于记载个体健康状况的个案数据(又称个人病历病案数据),所述个案数据包括但不限于被诊断人的名称、病历病案文档的文档编号、诊断日期、疾病名称、疾病编码等。
作为优选,服务器在接收到包含各测试组的数据调取指令时,自数据库中获取与各测试组关联的测试对象表,所述测试对象表中包含同一测试组中每一个测试对象的对象数据。所述服务器可以应用于寿命预估系统。
S20,接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数。
在本实施例中,所述数据匹配指令是指用户在客户端上传本地存储的指定项文件或者输入指定项标识,并点击与数据匹配指令绑定的预设按钮之后发送至服务器的;所述指定项文件中包含预设数量的指定项、对应于指定项的指定项标识、对应于指定项的调整权重等;所述指定项标识包括但不限于指定项的项目名称、项目编码或者项目关键词。
作为优选,在获取到用户在客户端上传的指定项文件时,预设的应用程序读取该指定项文件中的所有指定项标识之后,将所有指定项标识发送至服务器中,此时,服务器接收到包含指定项标识的数据匹配指令,利用双指标匹配法同步自每一个对象数据表中查询与指定项标识匹配的对象数据的个数,并根据与指定项标识匹配的对象数据的个数确定各对象数据表中有效数据个数。其中,所述双指标匹配法是指,利用正则匹配方式分别对象数据表中每一个对象数据中包含的名称或编码进行匹配。
S30,根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数。
在本实施例中,所述指定项损失系数用于计算每一类所述测试对象的损失寿命。
作为优选,将与指定项标识匹配的有效数据个数作为输入参数,输入至与指定项关联的系数模型(也即一个指标项关联一个系数模型),并接收该系数模型输出的指定项损失系数。其中,所述系数模型为:
Figure PCTCN2020093362-appb-000001
其中,Fx(i)为所述指定项损失系数;Ex(i)为所述指定项标识匹配的有效数据个数;w(i)为所述指定项对应的调整权重,在该系数模型中为常数;Mx为所述对象数据表的数据量。可理解的,根据上述系数模型中的
Figure PCTCN2020093362-appb-000002
可以获得指定项概率。
可理解的,在数据匹配指令中包含一个或多个指定项标识时,利用上述步骤S20和S30确定每一类测试对象中的对应于每一个指定项标识的指定项损失系数。
S40,接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命。
在本实施例中,所述数据分析指令是指用户在客户端上传存储在本地的寿命文件或者输入每一类测试对象的初始参数,并点击与数据分析指令绑定的预设按钮之后发送至服务器的;所述寿命文件中包含每一类测试对象的初始对象数量、被标记对象数量、被标记率、平均寿命、总寿命、初始预测寿命等初始参数;所述被标记率是指已经进入确切测试组的一类测试对象在该测试组的测试时长内被标记的可能性;所述初始对象数量是指刚进入某一测试组的一类测试对象未被标记的测试对象数量;所述标记对象数量是指在某一测试组的一类测试对象在测试时长内被标记的测试对象数量;所述平均寿命是指进入确切测试组的一类测试对象在进入另一确切测试组之间可能未被标记的时间长度;所述总寿命是指进入确切测试组的一类测试对象,按照某一被标记率确定该类测试对象在未来可能未被标记的时间总长;所述初始预期寿命是指进入某一确切测试组的一类测试对象,按照某一被标记率确定该类测试对象在未来可能未被标记的平均时间长度。
作为优选,所述寿命文件为以表格式模型存储的数据表,该数据表可以是指获取一个初始对象数量(例如,10万),并根据每一类测试对象的被标记对象数量确定每一类测试对象的被标记率,进一步地构成表格式模型,直至初始对象数量与被标记数量的差值小于或等于预设阈值(例如,0)为止。可理解的,
示例性的,若测试对象为产品,且测试组为产品的使用周期,则该寿命文件中包含产品从生产至达到淘汰为止的在各使用周期的被标记率(也即产品被标记为淘汰的概率),在各使用周期的被标记率可以作为用户决定是否购买产品的重要参考依据,也可以反映某一测试点的产品被淘汰规律。
示例性的,若测试对象为个体,且测试组为个体的年龄段,则该数据表中包含个体从出生直至死亡为止的在各年龄段的被标记率(也即个体被标记为死亡的概率),在各年龄段的被标记率可以作为用户投保的重要依据,也可以反映某一测试点的个体健康状况以及生存死亡规律。
且作为优选,在获取到用户在客户端上传的寿命文件时,预设的应用程序读取该寿命文件中的每一类测试对象的初始参数之后,将每一类测试对象的初始参数发送至服务器中,此时,服务器接收到包含每一类测试对象的初始参数的数据分析指令,将每一类测试对象的初始参数和上述步骤S30中的指定项损失系数组作为输入参数,输入至预设的寿命损失分析模 型中,并接收该寿命损失分析模型输出的每一类测试对象的输出的累积损失寿命,进而将每一类测试对象与对应的累积损失寿命关联存储至数据库中。
且作为优选,首先将每一类测试对象的指定项损失系数置于系数数组中,将每一类测试对象的初始参数置于参数数组中,将系数数组和参数数组一同输入至寿命损失分析模型中,此时,该寿命损失分析模型利用系数数组和参数数组中的数据进行迭代、累积等处理,得到包含每一类测试对象的累积损失寿命的输出数组,以提高数据处理效率,避免数据出错。
S50,将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
在本实施例中,所述调整模型为:
Adj·ex=ex-Total·Lx
其中,Adj·ex为所述调整预期寿命;ex为所述初始预期寿命;Total·Lx为累积损失寿命。
作为优选,将初始参数中包含的初始预期寿命和上述步骤S40中的累积损失寿命作为输入参数,输入至上述调整模型中,并接收上述调整模型输出的调整预期寿命,进而将每一类测试对象与对应的调整预期寿命关联存储至数据库中,此时,每一类测试对象的累积损失寿命与调整预期寿命亦进行关联存储。
进一步地,在一实施例中,在获取到每一类所述测试对象的调整预期寿命时,检测所述测试对象的调整预期寿命是否达到告警条件,在所述测试对象的调整预期寿命是否达到告警条件时,向所述客户端发出预警信息。
综上所述,本实施例根据数据调取指令自数据库获取各测试组的对象数据表,根据数据匹配指令自各对象数据表中统计与指定项标识匹配的有效数据个数,从而确定每一类测试对象的指定项损失系数,此时,根据数据分析指令将初始参数和指定项损失系数输入至寿命损失分析模型中,接收寿命损失分析模型输出的累积损失寿命,进而根据每一类测试对象的初始预期寿命和累积损失寿命,确定调整预期寿命,本实施例形成系统可复用的寿命预估方法,节约了成本,且使用的测试组数据更为丰富,提高了寿命预估的细粒度,且在寿命预估过程中提高了数据匹配效率、寿命预估精准度和寿命预估效率,且有利于后续寿命监测;同时客观反映了测试对象的寿命情况。
在一实施例中,所述的测试对象的寿命预估方法,还包括以下步骤:
首先,接收自客户端发送的包含用户标识的登录请求,在登录用户的身份验证通过之后,根据所述用户标识自预设的用户权限表中获取所述登录用户的查询权限等级。
然后,接收包含所述查询权限等级和查询条件的查询指令,获取与所述查询权限等级匹配且满足所述查询条件的查询数据,将所述查询数据以预设文档形式导出并发送至所述客户端;所述查询数据包括每一类所述测试对象的调整预期寿命。
在本实施例中所述用户权限表中包含用户标识、对应于用户标识的查询权限等级;所述用户标识包括但不限于用户名、用户编号;所述查询权限等级根据需求设置,例如:1~5级。
所述查询条件包括查询对象所在的测试点、查询对象所在的测试组、查询对象类别等。优选地,各测试点和对象数据表已关联存储至数据库中。
具体地,接收自客户端发送的登录指令,该登录指令中包含登录用户的用户标识,自用户表(包含一个或多个有效用户标识,且已预先存储在数据库)中查询是否存在与用户标识匹配的有效用户标识,在用户表中存在与用户标识匹配的有效用户标识时,确定登录用户的身份验证通过;而在用户表中不存在与用户标识匹配的有效用户标识时,确定登录用户的身份验证未通过。
进一步地,在登录用户的身份验证通过之后,自用户权限表中查询与用户标识对应的查询权限等级,并提示登录用户输入查询条件,进而在登录用户在客户端输入查询条件,并点 击与查询指令绑定的预设按钮时,服务器接收到包含查询条件和查询权限等级的查询指令,自数据库中查找与查询权限等级关联且满足查询条件的查询数据,并将查询数据、查询时间、查询条件、查询权限等级等添加至预设的数据表模板中生成查询数据表,并将查询数据表导出并发送至所述客户端,便于用户查看、编辑、分析查询数据。
综上所述,本实施例在接收并验证通过登录用户发送的登录请求,自用户权限表获得登录用户的查询权限等级,进而根据查询指令,查询并获取与查询权限等级匹配且满足查询条件的查询数据,避免了登录用户跨权限等级进行操作,保证了数据安全性和可靠性。
在一实施例中,如图3所示,所述步骤S10之前具体包括以下步骤:
S101,自预设的数据中心获取各所述测试组的测试对象集合,一个所述测试对象集合中包含同一类的至少一个测试对象,且每一个所述测试对象包含至少一个对象文档。
在本实施例中,所述对象文档可以是指记载对象数据,且储存在数据中心的纸质扫描件等,且所述对象文档可以是以标签图像文件格式、图片格式、便携式文档格式存储。
具体的,在接收数据调取指令之前,根据授权信息访问预设的数据中心(例如,针对测试对象为产品设置的产品检修中心、针对测试对象为个体设置的电子病历中心等),并根据调用信息自数据中心获取各测试组的测试对象集合,优选地,每一个测试对象集合中包含的测试对象数量和测试文档数量均相同;所述授权信息包括数据中心的授权编号;所述调取信息包括选取对象文档的数量、选取对象文档的顺序(例如,按照对象文档的存储时间选取)等,例如,随机选取与某一测试组关联的对象文档10000份。
S102,将同一个所述测试对象集合中的所有所述对象文档,分配至同一条数据提取线程中。也即,一条数据提取线程用于对一个测试对象集合中的所有对象文档进行识别,进而提取对象文档中包含的对象数据,可理解的,通过多条数据提取线程同步识别并提取数据,便于数据存储,且提高了数据处理效率。
S103,调用与所述数据提取线程关联的光学字符识别模型,对分配至所述数据提取线程中的所有所述对象文档进行识别,获取所述数据提取线程中每一个所述对象文档中包含的对象数据。也即,一条数据提取线程关联一个光学字符识别模型,利用该光学字符识别模型对分配至该数据提取线程中的所有对象文档进行识别,进而提取每一个对象文档中包含的对象数据,该对象数据包括但不限于对象文档中的被检测对象的基本信息、文档编号、文档日期、文档来源信息、被检测项。在本实施例中,通过光学字符识别模型对对象文档进行识别,时效更快、差错率更低;且该识别过程无需人工进行操作,不会产生人工操作需要停歇的时间间断,且同时可对多个对象文档进行识别,识别效率更高。
作为优选,所述光学识别字符识别模型是根据待识别文档训练生成的学习模型,此时,所述步骤S103之前包括以下步骤:获取待识别文档,并根据所述待识别文档训练生成光学字符识别模型。可理解的,所述待识别文档为存储在数据中心的历史对象文档,在训练所述光学字符识别模型时,可以使用2000张同类型的对象文档作为待识别文档,在每一次根据一张对象文档进行学习之后,均需要根据学习内容校正结果,在经过反复学习之后,生成可以对象文档的光学字符识别模型。
S104,根据所述数据提取线程中每一个所述对象文档中包含的对象数据生成对象数据表,并将所述对象数据表与对应的所述测试组关联存储至所述数据库中。也即,将同一数据提取线程提取到的所有对象数据添加至一个数据表模型中,从而生成对象数据表,并将对象数据表与对应的测试组关联存储至数据库中。可理解的,所述对象数据表可以根据用户需求进行更新。
综上所述,本实施例获取多个测试对象集合,并通过多条数据提取线程分别对各测试对象集合中的所有对象文档进行同步识别,便于数据存储,提高了识别效率,且多条提取线程各自进行识别,避免了数据相互干涉;且通过光学字符识别模型识别对象文档,时效更快,差错率更低,该整个识别过程无需时间间隔,进一步提高了识别效率;同时测试组数据丰富,提高了寿命预估的细粒度。
在一实施例中,如图4所示,所述步骤S20具体包括以下步骤:
S201,接收数据匹配指令,所述数据匹配指令包含各指定项的项目名称和项目编码。
S202,根据各所述指定项的项目名称,获取对应的名称匹配表达式。
S203,根据各所述指定项的项目编码,获取对应的编码匹配表达式。
S204,自各所述对象数据表中,获取并标记满足所述名称匹配表达式和/或所述编码匹配表达的所述对象数据,统计已标记的所述对象数据获得与各所述项目名称和/或各所述项目编码指匹配的有效数据个数。
在本实施例中,接收到的数据匹配指令还可以包含指定项的项目关键词等,通过上述步骤S202中的名称匹配表达式与上述步骤S203中的编码匹配表达式这两种正则匹配方式,最终在对象数据表中,查询并标记出匹配成功的对象数据,并将匹配成功的对象数据记录为有效对象数据;可理解的,两种匹配方式只要有一种满足要求,即为匹配成功。
示例性的,若测试对象为个体,获取的指定项标识(也即,指定疾病的疾病名称或关键词)为“中风”,使用“中风”在国家疾病分类标准编码中进行搜索,首先锁定疾病标准编码前三位在I60-I69的全部疾病,部分示例如表1所示,再通过对表1中的部分疾病名称进行总结,如都包含“梗死”、“梗塞”或者“偏瘫”等关键词,形成对疾病名称的文字匹配总结。
进一步地,根据需求将中风分为首次中风及长期中风患者,在疾病标准编码中前三位为I69对应所有的中风后遗症(如表2如下),同时对名称匹配表达式也做了相应的修改;
表1
疾病标准编码 疾病名称 助记码
I63.301 血栓性偏瘫 XSXPT
I63.351 脑软化伴脑血栓 NRHBNXS
I63.401 脑栓塞性偏瘫 NSSXPT
I63.801 脑干梗塞 NGGS
I63.851 动脉硬化性脑软化 DMYHXNRH
I63.852 脑血管病性脑软化 NXGBXNRH
I63.901 多发性脑梗塞 DFXNGS
I63.902 脑梗塞 NGS
表2
疾病标准编码 疾病名称 助记码
I69.051 蛛网膜下出血后遗症 ZWMXCXHYZ
I69.101 脑出血后遗症 NCXHYZ
I69.301 脑梗塞后遗症 NGSHYZ
I69.451 出血或梗死中风后遗症 CXHGSZFHYZ
I69.452 脑卒中后遗症 NZZHYZ
I69.801 脑血管病后遗症 NXGBHYZ
I69.802 脑血管病恢复期 NXGBHFQ
I69.803 缺血缺氧性脑后遗症 QXQYZNBHYZ
I69.851 脑血栓后遗症 NXSHYZ
最终形成的匹配表达式如表3所示:
表3
Figure PCTCN2020093362-appb-000003
可理解的,根据表3中的名称匹配表达式和/或编码匹配表达式自对象数据表中,查询并标记有相应确诊记录的数据。综上所述,本实施例通过双指标匹配方式设定对象数据的匹配标准,精准确定需要被标记的测试对象,达到了数据精准匹配、数据高效匹配的目的。
在一实施例中,所述步骤S30具体包括以下步骤:
首先,获取与所述指定项标识关联的调整权重。也即,已存储在数据库中的指定项文件中,每一个指定项对应有一个指定项标识和疾病权重,此时,根据指定项标识自指定项文件确定调整权重即可。
然后,获取各所述测试数据表的数据量。也即,各对象数据表中所有对象数据的个数。
最后,根据各所述测试数据表的数据量、与各所述指定项标识匹配的有效数据个数以及关联的调整权重,获取每一类所述测试对象的指定项损失系数。作为优选,在数据匹配指令中包含一个和多个指定项标识时,可以根据上述步骤中的调整权重和对象总数确定与各指定项标识对应的系数模型(参考所述步骤S30),将与各指定项标识匹配的有效数据个数输入至与各指定项标识对应的系数模型中,即可获取每一类测试对象的各指定项损失系数。综上所述,本实施例基于指定项损失系数调整初始预期寿命,有利于客观反映测试对象在各测试点的寿命情况。
在一实施例中,如图5所示,所述步骤S40具体包括以下步骤:
S401,接收数据分析指令,所述数据分析指令包含每一类所述测试对象的初始预期寿命、初始对象数量和总寿命。
在本实施例中,所述数据分析指令还可以包含被标记对象数量、被标记率、平均寿命等,其中,每一类测试对象的初始对象数量、总寿命以及初始预期寿命存在的关系如下:
Figure PCTCN2020093362-appb-000004
其中,Tx为所述总寿命;Ix为所述初始对象数量。
而每一类测试对象的初始对象数量、被标记对象数量以及被标记率存在的关系如下:
Figure PCTCN2020093362-appb-000005
其中,Mr为所述被标记率;Mx为所述被标记对象数量。
S402,通过所述寿命损失分析模型对每一类所述测试对象的所述初始对象数量、所述总寿命、各所述指定项损失系数进行迭代,得到每一类所述测试对象的指定项损失寿命。
作为优选,利用寿命损失分析模型中的迭代子模型,获得每一类所述测试对象的定项损失寿命的步骤如下:
首先,根据与最高测试组对应的一类测试对象的指定项损失系数和总寿命,获取与最高测试组对应的一类测试对象的损失总寿命,此时若令最高测试组为x=w-1,则有SUM(F w-1(i)*T w-1)=F w-1(i)*T w-1,其中SUM(F w-1(i)*T w-1)为与最高测试组对应的一类测试对象的损失总寿命,F w-1(i)为与最高测试组对应的一类测试对象的指定项损失系数,T w-1为与最高测试组对应的一类测试对象的总寿命。
然后,依次从高到低进行反向迭代,直至得到与第一测试组对应的一类测试对象的损失总寿命,即可得到每一类所述测试对象的损失总寿命,此时若令第一测试组为x=0,则有SUM(F 0(i)*T 0)=SUM(F 1(i)*T 1)+F 0(i)*T 0,其中SUM(F 0(i)*T 0)为与第一测试组对应的一类测试对象的损失总寿命,F 0(i)为与第一测试组对应的一类测试对象的指定项损失系数,T 0为与第一测试组对应的一类测试对象的总寿命;SUM(F 1(i)*T 1)为与第二测试组对应的一类测试对象的损失总寿命,F 1(i)为与第二测试组对应的一类测试对象的指定项损失系数,T 1为第二测试组对应的一类测试对象的总寿命。可理解的,第一测试组为初始化测试组。
最后,根据每一类测试对象的损失总寿命和初始对象数量,获得每一类测试对象的指定项损失寿命,可选地,利用计算公式
Figure PCTCN2020093362-appb-000006
可以获得每一类测试对象的指定项损失寿命,其中Lx(i)为每一类所述测试对象的指定项损失寿命;SUM(Fx(i)*Tx)为每一类所述测试对象的损失总寿命。
S403,通过所述寿命损失分析模型对每一类所述测试对象的指定项损失寿命进行累积,得到每一类所述测试对象的累积损失寿命。
作为优选,利用寿命损失分析模型中的累积子模型对所有指标项损失寿命(一个指定项对应一个指定项损失寿命)进行累积,获得累积损失寿命。其中,所述累积子模型为:
Figure PCTCN2020093362-appb-000007
其中,n为所述指定项数量。可理解的,在指定项为一个时,指定项损失寿命将作为累积损失寿命。综上所述,本实施例通过寿命损失分析模获得累积损失寿命,自动计算,提高了数据处理效率,同时提高了损失寿命的精准度。
在一实施例中,如图6所示,所述步骤S401之前,具体包括以下步骤:
S404,根据获取的每一类所述测试对象的测试对象数量和被标记对象数量,确定每一类所述测试对象的被标记率。在本实施例中,已知测试对象数量、被标记对象数量以及被标记率存在的关系(参考所述步骤S401)。
S405、将获取的与初始化测试组对应的一类所述测试对象的初始对象数量、每一类所述测试对象的所述被标记率输入至预设的正向迭代模型中,并接收所述正向迭代模型输出的每一类所述测试对象的所述初始对象数量。
在本实施例中,所述正向迭代模型为:
I j-1=I j-2-I j-2*Mr j-2
其中,I j-1为与下一个测试组对应的一类测试对象的初始对象数量,I j-2为与上一个测试组对应的一类测试对象的初始对象数量,Mr j-2为与上一个测试组对应的一类测试对象的被标记率。例如,当j-1=1时,可以得到I 1=I 0-I 0*Mr 0
S406、将每一类所述测试对象的所述初始对象数量输入至预设的反向迭代模型中,并接收所述反向迭代模型输出的每一类所述测试对象的总寿命。
在本实施例中,利用所述反向迭代模型获取每一类所述测试对象的总寿命的步骤如下:
首先,获取与最高测试组对应的一类测试对象的总寿命,令T w-1=∑I w-1*Q;其中,T w-1为与最高测试组对应的一类测试对象的总寿命,I w-1为与最高测试组对应的一类测试对象的初始对象数量,Q表示测试组的周期时长,例如1年。
然后,依次从高到低进行迭代计算,直至得到T 0=∑I 0*Q+T 1,即可得到每一类测试对象的总寿命。
S407、根据每一类所述测试对象的所述初始对象数量和所述总寿命,得到每一类所述测试对象的初始预期寿命,并将所述初始对象数量、所述总寿命和所述初始预期寿命关联存储至所述数据库中。
在本实施例中,已知每一类测试对象的初始对象数量、总寿命以及初始预期寿命存在的关系(参考所述步骤S401),优选地,可以根据每一类测试对象的所述初始对象数量、所述总寿命和所述初始预期寿命生成寿命文件,并将寿命文件存储至数据库中,后续步骤S401中,服务器自数据库获取寿命文件即可。综上所述,本实施例获取的初始参数丰富度高、准确度高,有利于提高寿命预估的精准度。
在一实施例中,如图7所示,提供一种测试对象的寿命预估装置,该测试对象的寿命预估装置与上述实施例中测试对象的寿命预估方法一一对应。该测试对象的寿命预估装置包括以下模块,各功能模块详细说明如下:
数据调取模块110,用于接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象。
数据匹配模块120,用于接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数。
数据确定模块130,用于根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数。
数据分析模块140,用于接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命。
数据调整模块150,用于将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
在一实施例中,所述的测试对象的寿命预估装置还包括以下模块,各功能模块详细说明如下:
登录模块,用于接收自客户端发送的包含用户标识的登录请求,在登录用户的身份验证通过之后,根据所述用户标识自预设的用户权限表中获取所述登录用户的查询权限等级。
查询显示模型,用于接收包含所述查询权限等级和查询条件的查询指令,获取与所述查询权限等级匹配且满足所述查询条件的查询数据,将所述查询数据以预设文档形式导出并发送至所述客户端;所述查询数据包括每一类所述测试对象的调整预期寿命。
在一实施例中,所述数据调取模块110包括以下子模块,各功能子模块详细说明如下:
调取子模块,用于自预设的数据中心获取各所述测试组的测试对象集合,一个所述测试对象集合中包含同一类的至少一个测试对象,且每一个所述测试对象包含至少一个对象文档。
分配子模块,用于将同一个所述测试对象集合中的所有所述对象文档,分配至同一条数据提取线程中。
调用子模块,用于调用与所述数据提取线程关联的光学字符识别模型,对分配至所述数据提取线程中的所有所述对象文档进行识别,获取所述数据提取线程中每一个所述对象文档中包含的对象数据。
第一存储子模块,用于根据所述数据提取线程中每一个所述对象文档中包含的对象数据生成对象数据表,并将所述对象数据表与对应的所述测试组关联存储至所述数据库中。
在一实施例中,所述数据匹配模块120包括以下子模块,各功能子子模块详细说明如下:
匹配子模块,用于接收数据匹配指令,所述数据匹配指令包含各指定项的项目名称和项目编码。
名称子模块,用于根据各所述指定项的项目名称,获取对应的名称匹配表达式。
编码子模块,用于根据各所述指定项的项目编码,获取对应的编码匹配表达式。
标记子模块,用于自各所述对象数据表中,获取并标记满足所述名称匹配表达式和/或所述编码匹配表达的所述对象数据,统计已标记的所述对象数据获得与各所述项目名称和/或各所述项目编码指匹配的有效数据个数。
在一实施例中,所述数据确定模块130包括以下子模块,各功能子子模块详细说明如下:
第一获取子模块,用于获取与所述指定项标识关联的调整权重。
第二获取子模块,用于获取各所述测试数据表的数据量。
第三获取子模块,用于根据各所述测试数据表的数据量、与各所述指定项标识匹配的有效数据个数以及关联的调整权重,获取每一类所述测试对象的指定项损失系数。
在一实施例中,所述数据分析模块140包括以下子模块,各功能子子模块详细说明如下:
分析子模块,用于接收数据分析指令,所述数据分析指令包含每一类所述测试对象的初始预期寿命、初始对象数量和总寿命。
迭代子模块,用于通过所述寿命损失分析模型对每一类所述测试对象的所述初始对象数量、所述总寿命、各所述指定项损失系数进行迭代,得到每一类所述测试对象的指定项损失寿命。
累积子模块,用于通过所述寿命损失分析模型对每一类所述测试对象的指定项损失寿命进行累积,得到每一类所述测试对象的累积损失寿命。
在另一实施例中,所述数据分析模块140还包括以下子模块,各功能子模块详细说明如下:
第四获取子模块,用于根据获取的每一类所述测试对象的测试对象数量和被标记对象数量,确定每一类所述测试对象的被标记率;
正向迭代子模块,用于将获取的与初始化测试组对应的一类所述测试对象的初始对象数量、每一类所述测试对象的所述被标记率输入至预设的正向迭代模型中,并接收所述正向迭代模型输出的每一类所述测试对象的所述初始对象数量;
反向迭代子模块,用于将每一类所述测试对象的所述初始对象数量输入至预设的反向迭 代模型中,并接收所述反向迭代模型输出的每一类所述测试对象的总寿命;
第二存储子模块,用于根据每一类所述测试对象的所述初始对象数量和所述总寿命,得到每一类所述测试对象的初始预期寿命,并将所述初始对象数量、所述总寿命和所述初始预期寿命关联存储至所述数据库中。
关于测试对象的寿命预估装置的具体限定可以参见上文中对于测试对象的寿命预估方法的限定,在此不再赘述。上述测试对象的寿命预估装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。上述各模块可以硬件形式内嵌于或独立于计算机设备中的处理器中,也可以以软件形式存储于计算机设备中的存储器中,以便于处理器调用执行以上各个模块对应的操作。
在一个实施例中,提供了一种计算机设备,该计算机设备可以是服务器,其内部结构图可以如图8所示。该计算机设备包括通过系统总线连接的处理器、存储器、网络接口和数据库。其中,该计算机设备的处理器用于提供计算和控制能力。该计算机设备的存储器包括可读存储介质、内存储器。该可读存储介质存储有操作系统、计算机可读指令和数据库。该内存储器为可读存储介质中的操作系统和计算机可读指令的运行提供环境。该计算机可读指令被处理器执行时以实现一种测试对象的寿命预估方法。本实施例所提供的可读存储介质包括非易失性可读存储介质和易失性可读存储介质。
在一个实施例中,提供了一种计算机设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机可读指令,处理器执行计算机可读指令时实现以下步骤:
接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
在一个实施例中,提供了一个或多个存储有计算机可读指令的可读存储介质,本实施例所提供的可读存储介质包括非易失性可读存储介质和易失性可读存储介质;该可读存储介质上存储有计算机可读指令,该计算机可读指令被一个或多个处理器执行时,使得一个或多个处理器实现以下步骤:
接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机可读指令来指令相关的硬件来完成,所述的计算机可读指令可存储于一非易失性计算机 可读取存储介质或易失性可读存储介质中,该计算机可读指令在执行时,可包括如上述各方法的实施例的流程。其中,本申请所提供的各实施例中所使用的对存储器、存储、数据库或其它介质的任何引用,均可包括非易失性和/或易失性存储器。非易失性存储器可包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双数据率SDRAM(DDRSDRAM)、增强型SDRAM(ESDRAM)、同步链路DRAM(SLDRAM)、存储器总线直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。
所属领域的技术人员可以清楚地了解到,为了描述的方便和简洁,仅以上述各功能单元或模块的划分进行举例说明,实际应用中,可以根据需要而将上述功能分配由不同的功能单元或模块完成,即将所述装置的内部结构划分成不同的功能单元或模块,以完成以上描述的全部或者部分功能。
以上所述实施例仅用以说明本申请的技术方案,而非对其限制;尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围,均应包含在本申请的保护范围之内。

Claims (20)

  1. 一种测试对象的寿命预估方法,其中,包括:
    接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
    接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
    根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
    接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
    将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
  2. 如权利要求1所述的测试对象的寿命预估方法,其中,所述接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表之前,包括:
    自预设的数据中心获取各所述测试组的测试对象集合,一个所述测试对象集合中包含同一类的至少一个测试对象,且每一个所述测试对象包含至少一个对象文档;
    将同一个所述测试对象集合中的所有所述对象文档,分配至同一条数据提取线程中;
    调用与所述数据提取线程关联的光学字符识别模型,对分配至所述数据提取线程中的所有所述对象文档进行识别,获取所述数据提取线程中每一个所述对象文档中包含的对象数据;
    根据所述数据提取线程中每一个所述对象文档中包含的对象数据生成对象数据表,并将所述对象数据表与对应的所述测试组关联存储至所述数据库中。
  3. 如权利要求1所述的测试对象的寿命预估方法,其中,所述接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数,包括:
    接收数据匹配指令,所述数据匹配指令包含各指定项的项目名称和项目编码;
    根据各所述指定项的项目名称,获取对应的名称匹配表达式;
    根据各所述指定项的项目编码,获取对应的编码匹配表达式;
    自各所述对象数据表中,获取并标记满足所述名称匹配表达式和/或所述编码匹配表达的所述对象数据,统计已标记的所述对象数据获得与各所述项目名称和/或各所述项目编码指匹配的有效数据个数。
  4. 如权利要求1所述的测试对象的寿命预估方法,其中,所述根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数,包括:
    获取与所述指定项标识关联的调整权重;
    获取各所述测试数据表的数据量;
    根据各所述测试数据表的数据量、与各所述指定项标识匹配的有效数据个数以及关联的调整权重,获取每一类所述测试对象的指定项损失系数。
  5. 如权利要求1所述的测试对象的寿命预估方法,其中,所述接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命,包括:
    接收数据分析指令,所述数据分析指令包含每一类所述测试对象的初始预期寿命、初始对象数量和总寿命;
    通过所述寿命损失分析模型对每一类所述测试对象的所述初始对象数量、所述总寿命、各所述指定项损失系数进行迭代,得到每一类所述测试对象的指定项损失寿命;
    通过所述寿命损失分析模型对每一类所述测试对象的指定项损失寿命进行累积,得到每 一类所述测试对象的累积损失寿命。
  6. 如权利要求1所述的测试对象的寿命预估方法,其中,所述将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命之后,包括:
    接收自客户端发送的包含用户标识的登录请求,在登录用户的身份验证通过之后,根据所述用户标识自预设的用户权限表中获取所述登录用户的查询权限等级;
    接收包含所述查询权限等级和查询条件的查询指令,获取与所述查询权限等级匹配且满足所述查询条件的查询数据,将所述查询数据以预设文档形式导出并发送至所述客户端;所述查询数据包括每一类所述测试对象的调整预期寿命。
  7. 如权利要求1所述的测试对象的寿命预估方法,其中,所述接收包含每一类所述测试对象的初始参数的数据分析指令之前,包括:
    根据获取的每一类所述测试对象的测试对象数量和被标记对象数量,确定每一类所述测试对象的被标记率;
    将获取的与初始化测试组对应的一类所述测试对象的初始对象数量、每一类所述测试对象的所述被标记率输入至预设的正向迭代模型中,并接收正向迭代模型输出的每一类所述测试对象的所述初始对象数量;
    将每一类所述测试对象的所述初始对象数量输入至预设的反向迭代模型中,并接收所述反向迭代模型输出的每一类所述测试对象的总寿命;
    根据每一类所述测试对象的所述初始对象数量和所述总寿命,得到每一类所述测试对象的初始预期寿命,并将所述初始对象数量、所述总寿命和所述初始预期寿命关联存储至所述数据库中。
  8. 一种测试对象的寿命预估装置,其中,包括:
    数据调取模块,用于接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
    数据匹配模块,用于接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
    数据确定模块,用于根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
    数据分析模块,用于接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命;
    数据调整模块,用于将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
  9. 如权利要求8所述的测试对象的寿命预估装置,其中,所述数据调取模块包括:
    调取子模块,用于自预设的数据中心获取各所述测试组的测试对象集合,一个所述测试对象集合中包含同一类的至少一个测试对象,且每一个所述测试对象包含至少一个对象文档;
    分配子模块,用于将同一个所述测试对象集合中的所有所述对象文档,分配至同一条数据提取线程中;
    调用子模块,用于调用与所述数据提取线程关联的光学字符识别模型,对分配至所述数据提取线程中的所有所述对象文档进行识别,获取所述数据提取线程中每一个所述对象文档中包含的对象数据;
    第一存储子模块,用于根据所述数据提取线程中每一个所述对象文档中包含的对象数据生成对象数据表,并将所述对象数据表与对应的所述测试组关联存储至所述数据库中。
  10. 如权利要求8所述的测试对象的寿命预估装置,其中,所所述数据匹配模块包括:
    匹配子模块,用于接收数据匹配指令,所述数据匹配指令包含各指定项的项目名称和项目编码;
    名称子模块,用于根据各所述指定项的项目名称,获取对应的名称匹配表达式;
    编码子模块,用于根据各所述指定项的项目编码,获取对应的编码匹配表达式;
    标记子模块,用于自各所述对象数据表中,获取并标记满足所述名称匹配表达式和/或所述编码匹配表达的所述对象数据,统计已标记的所述对象数据获得与各所述项目名称和/或各所述项目编码指匹配的有效数据个数。
  11. 如权利要求8所述的测试对象的寿命预估装置,其中,所所述数据确定模块包括:
    第一获取子模块,用于获取与所述指定项标识关联的调整权重;
    第二获取子模块,用于获取各所述测试数据表的数据量;
    第三获取子模块,用于根据各所述测试数据表的数据量、与各所述指定项标识匹配的有效数据个数以及关联的调整权重,获取每一类所述测试对象的指定项损失系数。
  12. 如权利要求8所述的测试对象的寿命预估装置,其中,所所述数据分析模块包括:
    分析子模块,用于接收数据分析指令,所述数据分析指令包含每一类所述测试对象的初始预期寿命、初始对象数量和总寿命;
    迭代子模块,用于通过所述寿命损失分析模型对每一类所述测试对象的所述初始对象数量、所述总寿命、各所述指定项损失系数进行迭代,得到每一类所述测试对象的指定项损失寿命;
    累积子模块,用于通过所述寿命损失分析模型对每一类所述测试对象的指定项损失寿命进行累积,得到每一类所述测试对象的累积损失寿命。
  13. 如权利要求8所述的测试对象的寿命预估装置,其中,所述测试对象的寿命预估装置还包括:
    登录模块,用于接收自客户端发送的包含用户标识的登录请求,在登录用户的身份验证通过之后,根据所述用户标识自预设的用户权限表中获取所述登录用户的查询权限等级;
    查询显示模型,用于接收包含所述查询权限等级和查询条件的查询指令,获取与所述查询权限等级匹配且满足所述查询条件的查询数据,将所述查询数据以预设文档形式导出并发送至所述客户端;所述查询数据包括每一类所述测试对象的调整预期寿命。
  14. 如权利要求8所述的测试对象的寿命预估装置,其中,所述数据分析模块还包括:
    第四获取子模块,用于根据获取的每一类所述测试对象的测试对象数量和被标记对象数量,确定每一类所述测试对象的被标记率;
    正向迭代子模块,用于将获取的与初始化测试组对应的一类所述测试对象的初始对象数量、每一类所述测试对象的所述被标记率输入至预设的正向迭代模型中,并接收所述正向迭代模型输出的每一类所述测试对象的所述初始对象数量;
    反向迭代子模块,用于将每一类所述测试对象的所述初始对象数量输入至预设的反向迭代模型中,并接收所述反向迭代模型输出的每一类所述测试对象的总寿命;
    第二存储子模块,用于根据每一类所述测试对象的所述初始对象数量和所述总寿命,得到每一类所述测试对象的初始预期寿命,并将所述初始对象数量、所述总寿命和所述初始预期寿命关联存储至所述数据库中。
  15. 一种计算机设备,包括存储器、处理器以及存储在所述存储器中并可在所述处理器上运行的计算机可读指令,其中,所述处理器执行所述计算机可读指令时实现如下步骤:
    接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
    接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
    根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
    接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每 一类所述测试对象的累积损失寿命;
    将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
  16. 如权利要求15所述的计算机设备,其中,所述接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表之前,所述处理器执行所述计算机可读指令时还实现如下步骤:
    自预设的数据中心获取各所述测试组的测试对象集合,一个所述测试对象集合中包含同一类的至少一个测试对象,且每一个所述测试对象包含至少一个对象文档;
    将同一个所述测试对象集合中的所有所述对象文档,分配至同一条数据提取线程中;
    调用与所述数据提取线程关联的光学字符识别模型,对分配至所述数据提取线程中的所有所述对象文档进行识别,获取所述数据提取线程中每一个所述对象文档中包含的对象数据;
    根据所述数据提取线程中每一个所述对象文档中包含的对象数据生成对象数据表,并将所述对象数据表与对应的所述测试组关联存储至所述数据库中。
  17. 如权利要求15所述的计算机设备,其中,所述接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数,包括:
    接收数据匹配指令,所述数据匹配指令包含各指定项的项目名称和项目编码;
    根据各所述指定项的项目名称,获取对应的名称匹配表达式;
    根据各所述指定项的项目编码,获取对应的编码匹配表达式;
    自各所述对象数据表中,获取并标记满足所述名称匹配表达式和/或所述编码匹配表达的所述对象数据,统计已标记的所述对象数据获得与各所述项目名称和/或各所述项目编码指匹配的有效数据个数。
  18. 如权利要求15所述的计算机设备,其中,所述根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数,包括:
    获取与所述指定项标识关联的调整权重;
    获取各所述测试数据表的数据量;
    根据各所述测试数据表的数据量、与各所述指定项标识匹配的有效数据个数以及关联的调整权重,获取每一类所述测试对象的指定项损失系数。
  19. 如权利要求15所述的计算机设备,其中,所述接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每一类所述测试对象的累积损失寿命,包括:
    接收数据分析指令,所述数据分析指令包含每一类所述测试对象的初始预期寿命、初始对象数量和总寿命;
    通过所述寿命损失分析模型对每一类所述测试对象的所述初始对象数量、所述总寿命、各所述指定项损失系数进行迭代,得到每一类所述测试对象的指定项损失寿命;
    通过所述寿命损失分析模型对每一类所述测试对象的指定项损失寿命进行累积,得到每一类所述测试对象的累积损失寿命。
  20. 一个或多个存储有计算机可读指令的可读存储介质,其中,所述计算机可读指令被一个或多个处理器执行时,使得所述一个或多个处理器执行如下步骤:
    接收包含各测试组的数据调取指令,自数据库获取与各所述测试组关联的对象数据表;一个所述测试组对应一类测试对象;
    接收包含指定项标识的数据匹配指令,自各所述对象数据表统计与所述指定项标识匹配的有效数据个数;
    根据与所述指定项标识匹配的有效数据个数,确定每一类所述测试对象的指定项损失系数;
    接收包含每一类所述测试对象的初始参数的数据分析指令,将各所述指定项损失系数和各所述初始参数输入至预设的寿命损失分析模型中,并接收所述寿命损失分析模型输出的每 一类所述测试对象的累积损失寿命;
    将每一类所述测试对象的初始预期寿命和所述累积损失寿命输入至预设的调整模型中,接收所述调整模型输出的每一类所述测试对象的调整预期寿命。
PCT/CN2020/093362 2019-06-14 2020-05-29 测试对象的寿命预估方法、装置、设备及介质 WO2020248844A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910516569.6 2019-06-14
CN201910516569.6A CN110390424B (zh) 2019-06-14 2019-06-14 测试对象的寿命预估方法、装置、设备及介质

Publications (1)

Publication Number Publication Date
WO2020248844A1 true WO2020248844A1 (zh) 2020-12-17

Family

ID=68285366

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/093362 WO2020248844A1 (zh) 2019-06-14 2020-05-29 测试对象的寿命预估方法、装置、设备及介质

Country Status (2)

Country Link
CN (1) CN110390424B (zh)
WO (1) WO2020248844A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112862349A (zh) * 2021-03-04 2021-05-28 北京同邦卓益科技有限公司 基于abs业务数据的数据处理方法、装置和设备
CN117893244A (zh) * 2024-03-15 2024-04-16 中国海洋大学 一种基于机器学习的海藻水热炭化应用综合管控系统
CN117893244B (zh) * 2024-03-15 2024-06-04 中国海洋大学 一种基于机器学习的海藻水热炭化应用综合管控系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110390424B (zh) * 2019-06-14 2022-05-17 平安科技(深圳)有限公司 测试对象的寿命预估方法、装置、设备及介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105809343A (zh) * 2016-03-04 2016-07-27 深圳市前海安测信息技术有限公司 用于评估保险标的风险的核保和精算数据库系统
US20180150609A1 (en) * 2016-11-29 2018-05-31 Electronics And Telecommunications Research Institute Server and method for predicting future health trends through similar case cluster based prediction models
CN108549951A (zh) * 2018-03-09 2018-09-18 璇飞(武汉)科技有限公司 一种基于关键器件的船用电气设备寿命预测方法及装置
CN108765094A (zh) * 2018-06-06 2018-11-06 中国平安人寿保险股份有限公司 保险数据处理方法、装置、计算机设备和存储介质
CN109061333A (zh) * 2018-06-21 2018-12-21 国网木星(北京)科技有限公司 能源管理系统设备实际寿命测试方法及系统
CN110390424A (zh) * 2019-06-14 2019-10-29 平安科技(深圳)有限公司 测试对象的寿命预估方法、装置、设备及介质

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101447048B (zh) * 2008-12-30 2011-08-03 上海发电设备成套设计研究院 一种变压器绝缘寿命预测的方法与寿命管理系统
KR102343967B1 (ko) * 2014-12-04 2021-12-28 삼성전자주식회사 배터리의 상태를 추정하는 방법 및 장치
CN108613889B (zh) * 2018-04-27 2020-07-07 佛山科学技术学院 基于循环寿命的钛合金钝缺口疲劳强度损失系数评估方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105809343A (zh) * 2016-03-04 2016-07-27 深圳市前海安测信息技术有限公司 用于评估保险标的风险的核保和精算数据库系统
US20180150609A1 (en) * 2016-11-29 2018-05-31 Electronics And Telecommunications Research Institute Server and method for predicting future health trends through similar case cluster based prediction models
CN108549951A (zh) * 2018-03-09 2018-09-18 璇飞(武汉)科技有限公司 一种基于关键器件的船用电气设备寿命预测方法及装置
CN108765094A (zh) * 2018-06-06 2018-11-06 中国平安人寿保险股份有限公司 保险数据处理方法、装置、计算机设备和存储介质
CN109061333A (zh) * 2018-06-21 2018-12-21 国网木星(北京)科技有限公司 能源管理系统设备实际寿命测试方法及系统
CN110390424A (zh) * 2019-06-14 2019-10-29 平安科技(深圳)有限公司 测试对象的寿命预估方法、装置、设备及介质

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112862349A (zh) * 2021-03-04 2021-05-28 北京同邦卓益科技有限公司 基于abs业务数据的数据处理方法、装置和设备
CN112862349B (zh) * 2021-03-04 2024-04-16 北京同邦卓益科技有限公司 基于abs业务数据的数据处理方法、装置和设备
CN117893244A (zh) * 2024-03-15 2024-04-16 中国海洋大学 一种基于机器学习的海藻水热炭化应用综合管控系统
CN117893244B (zh) * 2024-03-15 2024-06-04 中国海洋大学 一种基于机器学习的海藻水热炭化应用综合管控系统

Also Published As

Publication number Publication date
CN110390424B (zh) 2022-05-17
CN110390424A (zh) 2019-10-29

Similar Documents

Publication Publication Date Title
US11822975B2 (en) Systems and methods for synthetic data generation for time-series data using data segments
US20230126005A1 (en) Consistent filtering of machine learning data
WO2020186786A1 (zh) 文件处理方法、装置、计算机设备和存储介质
US11182691B1 (en) Category-based sampling of machine learning data
WO2021047326A1 (zh) 信息推荐方法、装置、计算机设备和存储介质
WO2021012790A1 (zh) 页面数据生成方法、装置、计算机设备及存储介质
US8671112B2 (en) Methods and apparatus for automated image classification
CN111178069B (zh) 数据处理方法、装置、计算机设备及存储介质
WO2022062353A1 (zh) 医疗数据处理方法、装置、计算机设备和存储介质
WO2020248844A1 (zh) 测试对象的寿命预估方法、装置、设备及介质
US11474978B2 (en) Systems and methods for a data search engine based on data profiles
CN109492191B (zh) 计算投保概率的方法、装置、计算机设备和存储介质
CN109117440B (zh) 一种元数据信息获取方法、系统和计算机可读存储介质
US11347719B2 (en) Multi-table data validation tool
CN112035611B (zh) 目标用户推荐方法、装置、计算机设备和存储介质
WO2021012482A1 (zh) 群体兴趣标签的生成方法、装置、计算机设备和存储介质
CA2937919A1 (en) Evaluating data quality of clinical trials
CN113836131B (zh) 一种大数据清洗方法、装置、计算机设备及存储介质
CN109658126B (zh) 基于产品推广的数据处理方法、装置、设备及存储介质
CN112331349B (zh) 一种脑卒中复发监测系统
CN112035437A (zh) 病案数据的传输方法、装置、计算机设备及存储介质
WO2022134357A1 (zh) 分诊数据处理方法、装置、计算机设备及存储介质
CN110362607B (zh) 异常号码识别方法、装置、计算机设备及存储介质
WO2020233310A1 (zh) 实时处理数据的异步补偿方法、装置、设备及存储介质
CN111210884B (zh) 一种临床医疗数据采集方法、装置、介质及设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20823653

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20823653

Country of ref document: EP

Kind code of ref document: A1