CN113988369A - Prediction model training method and prediction method based on prediction model - Google Patents
Prediction model training method and prediction method based on prediction model Download PDFInfo
- Publication number
- CN113988369A CN113988369A CN202111117392.6A CN202111117392A CN113988369A CN 113988369 A CN113988369 A CN 113988369A CN 202111117392 A CN202111117392 A CN 202111117392A CN 113988369 A CN113988369 A CN 113988369A
- Authority
- CN
- China
- Prior art keywords
- prediction model
- target
- training
- sample data
- parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (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
The embodiment of the invention provides a training method of a prediction model and a prediction method based on the prediction model, wherein the training method comprises the following steps: constructing a first prediction model, wherein the first prediction model is obtained by training historical operation sample data of the operation machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target operation sample data based on the first prediction model; determining a first target parameter corresponding to the first prediction model based on the first training parameter; constructing a second prediction model based on target historical operation sample data and the first training parameter; determining a second target parameter corresponding to the second prediction model; and adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model. The invention is used for solving the defects of low efficiency and high energy consumption of the excavator during operation aiming at complex and changeable operation objects in the prior art.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a prediction model training method and a prediction method based on a prediction model.
Background
The excavator is used as an engineering machine with high bearing capacity and running on a complex ground working condition, the running environment and the operation object of the excavator are complex and changeable, the power output intensity is high, and the fuel consumption is very high. Therefore, it is important to optimize the operation state of the excavator according to the operation request of the excavator to improve the operation efficiency of the excavator and to reduce the fuel consumption of the excavator. Further, since the work object is complicated and varied, the operating state of the excavator cannot be maintained under a certain constant condition, for example, the travel speed of the excavator varies due to variations in the flatness of the ground of the work object, the excavation speed varies due to variations in the hardness of the ground of the work object, and the like.
Therefore, how to provide a corresponding operation strategy for the excavator in time when the excavator faces a complex and variable operation object, so that the excavator adapts to the current operation object, and improves the operation efficiency and reduces the fuel consumption while adapting to the operation object is an important issue to be solved in the industry at present.
Disclosure of Invention
The embodiment of the invention provides a training method of a prediction model and a prediction method based on the prediction model, which are used for solving the defects of low efficiency and high energy consumption of an excavator during operation aiming at complex and changeable operation objects in the prior art, and the operation efficiency is improved and the fuel consumption is reduced while the excavator is adaptive to the operation objects.
The embodiment of the invention provides a training method of a prediction model, which comprises the following steps:
constructing a first prediction model, wherein the first prediction model is obtained by training historical operation sample data of a working machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target historical operation sample data based on the first prediction model;
determining a first target parameter corresponding to the first prediction model based on the first training parameter;
constructing a second prediction model based on the target historical operation sample data and the first training parameter;
determining a second target parameter corresponding to the second prediction model;
and adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.
According to a training method of a prediction model of an embodiment of the present invention, determining target historical operation sample data and a first training parameter corresponding to the target historical operation sample data based on the first prediction model includes:
determining at least one peak and at least one valley of the first predictive model;
taking historical operation sample data corresponding to the at least one peak and the at least one valley as the target historical operation sample data;
and respectively inputting the target historical operation sample data into a first preset calculation formula, and obtaining the first training parameter through the first preset calculation formula.
According to the training method of the prediction model of an embodiment of the present invention, the determining a first target parameter corresponding to the first prediction model based on the first training parameter includes:
determining a first target value from the first training parameters based on a second preset calculation formula, wherein the first target value is used for indicating the robustness;
determining the first target parameter corresponding to the first target value based on the first prediction model.
According to the training method of the prediction model, the building of the second prediction model based on the target historical operation sample data and the first training parameter includes:
combining the target historical operation sample data to obtain a first set;
combining the first training parameters to obtain a second set;
fitting to obtain a mapping relation between the first set and the second set;
and constructing a second prediction model based on the target historical operation sample data, the first training parameter and the mapping relation.
According to a training method of a prediction model of an embodiment of the present invention, the determining a second target parameter corresponding to the second prediction model includes:
and taking the position corresponding to the maximum value as the second target parameter based on the second prediction model.
Before the training method of the prediction model according to an embodiment of the present invention, before adjusting the second prediction model based on the first target parameter and the second target parameter to obtain the final prediction model, the method further includes:
inputting the second target parameter into the first preset calculation formula, and obtaining a second target value through the first preset calculation formula;
adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model, including:
calculating a difference between the first target parameter and the second target parameter;
judging whether the difference value is within a preset range;
when the difference value is judged to be within the preset range, determining that the training of the prediction model is finished;
when the difference value is not determined to be within the preset range, adding the second target parameter to the first set, adding the second target value to the second set, and training the second prediction model based on the added first set and the added second set until the training of the second prediction model is completed;
when the difference is determined not to be within the preset range and the first target value is smaller than the second target value, updating the first target value to the second target value and the first target parameter to the second target parameter; training the second prediction model based on the updated first target value and the updated first target parameter until the training of the second prediction model is completed.
The embodiment of the invention also provides a prediction method based on the prediction model, which comprises the following steps:
acquiring operation data of an operation machine during operation;
and inputting the operation data into the prediction model, obtaining resource data output by the prediction model, and sending the resource data to terminal equipment so as to enable the terminal equipment to output the resource data.
The embodiment of the present invention further provides a training device for a prediction model, including:
the system comprises a first construction module, a second construction module and a third construction module, wherein the first construction module is used for constructing a first prediction model, and the first prediction model is obtained by training historical operation sample data of a working machine and resource sample data corresponding to the historical operation sample data;
the first determination module is used for determining target historical operation sample data and a first training parameter corresponding to the target historical operation sample data based on the first prediction model;
a second determining module, configured to determine, based on the first training parameter, a first target parameter corresponding to the first prediction model;
the second construction module is used for constructing a second prediction model based on the target historical operation sample data and the first training parameter;
a third determining module, configured to determine a second target parameter corresponding to the second prediction model;
and the adjusting module is used for adjusting the second prediction model based on the first target parameter and the second target parameter so as to obtain a final prediction model.
The embodiment of the present invention further provides a prediction device based on a prediction model, including:
the acquisition module is used for acquiring operation data of the operation machine during operation;
and the output module is used for inputting the operation data into the prediction model, obtaining the resource data output by the prediction model, and sending the resource data to the terminal equipment so as to enable the terminal equipment to output the resource data.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method for training the predictive model as described in any of the above, or the steps of the predictive model-based predictive method, when executing the program.
According to the training method of the prediction model and the prediction method based on the prediction model, provided by the embodiment of the invention, the first prediction model is constructed and is obtained by training historical operation sample data of the operation machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target operation sample data based on the first prediction model; determining a first target parameter corresponding to the first prediction model based on the first training parameter; constructing a second prediction model based on target historical operation sample data and the first training parameter; determining a second target parameter corresponding to the second prediction model; the method comprises the steps of adjusting a first prediction model based on a first target parameter and a second target parameter to obtain a final prediction model, wherein the first prediction model is adjusted according to the first target parameter and the second target parameter, and the second prediction model is adjusted according to the first target parameter and the second target parameter to complete the training of the prediction model.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic flow chart of a method for training a prediction model according to an embodiment of the present invention;
FIG. 2 is a second flowchart illustrating a method for training a prediction model according to an embodiment of the present invention;
FIG. 3 is a third flowchart illustrating a method for training a prediction model according to an embodiment of the present invention;
FIG. 4 is a fourth flowchart illustrating a method for training a prediction model according to an embodiment of the present invention;
FIG. 5 is a fifth flowchart illustrating a method for training a prediction model according to an embodiment of the present invention;
FIG. 6 is a sixth flowchart illustrating a method for training a prediction model according to an embodiment of the present invention;
FIG. 7 is a flow chart illustrating a prediction method according to an embodiment of the present invention;
FIG. 8 is a schematic structural diagram of a training apparatus for a prediction model according to an embodiment of the present invention;
FIG. 9 is a schematic structural diagram of a prediction apparatus according to an embodiment of the present invention;
fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The following describes a training method of a prediction model according to an embodiment of the present invention with reference to fig. 1 to 6.
The embodiment of the invention provides a method for training a prediction model, which can be applied to an intelligent terminal, such as a mobile phone, a computer, a tablet and the like, can also be applied to a server, and can also be applied to a controller of a working machine. The method is described below by using the server as an example, but the method is only described by way of example and is not intended to limit the scope of the present invention. The other descriptions in the embodiments of the present invention are also for illustration purposes, and are not intended to limit the scope of the present invention.
The specific implementation of the training method of the prediction model is specifically shown in fig. 1:
The first prediction model is obtained through training of historical operation sample data of the working machine and resource sample data corresponding to the historical operation sample data.
Wherein, the operation machinery includes: excavators, electric excavators, hydraulic excavators, loaders, and the like.
Specifically, the historical job sample data includes: the data include operation data of the excavator obtained by a sensor mounted on the excavator, and environment data of a work object of the excavator during work. Wherein the operational data includes: the excavation speed of the excavator, the transfer speed of the excavator, and the like, and the environmental data includes: the flatness of the ground of the work object, the hardness of the ground of the work object, and the like.
Specifically, the resource sample data includes: fuel consumption of the excavator, power consumption of the excavator, time consumption of the excavator, and the like.
Next, an operation scene in which the excavator performs an operation on a flat ground, historical operation sample data as a transfer speed of the excavator, and resource sample data as fuel consumption of the excavator will be described as an example. It should be noted that the present embodiment is only an example, and is not intended to limit the scope of the present invention.
Specifically, before making the detailed description, the present invention will be described with reference to the following definitions.
First, a first objective function and a second objective function are defined, the first objective function corresponding to the first prediction model, and the second objective function corresponding to the second prediction model. The function corresponding to the thin curve (first curve) in fig. 2 is a first objective function f (x), and the function corresponding to the thick curve (second curve) in fig. 2 is a second objective functionNumber feff(x) It should be noted that the thick curve and the thin curve in fig. 2 are only used to distinguish the two functions, and do not refer to other meanings.
The present invention minimizes fuel consumption of the excavator by finding the point of the position of the five-pointed star marker in the second curve. In the following, a robust solution is defined: corresponding to the maximized objective function f (x), a solution x*The precondition called a robust solution is x*Is a globally optimal solution of equation (1).
wherein x is [ x ]1,x2,…,xD]TIs a decision variable, i.e. the transfer speed of the excavator, omega is the feasible domain of the decision variable, f (x, delta) is the robust function value, and f (x, delta) can also be usedeff(x) It is shown that,is a delta neighborhood of x, delta ═ delta1,δ2,…,δD]TAnd representing the disturbance intensity of the decision variable, wherein H is the sampling number set by a user, and k is a variable.
Fig. 2 illustrates an example where δ is 1.
For example, if δ is 1, the specific calculation steps for the robust function value at a speed of 3.6 are:
determining oil consumption corresponding to a preset number of speeds within the interval of [ 2.6-4.6 ], wherein the preset number is 10000 for example. See in particular the dots in the dotted lined area of fig. 2. And calculating the average value of 10000 oil consumptions, and taking the obtained average value as a robust function value.
In a specific embodiment, the manner of obtaining the target historical operation sample data and the first training parameter is specifically as shown in fig. 3:
at step 301, at least one peak and at least one valley of the first predictive model are determined.
Specifically, the first prediction model is optimized by using an optimization algorithm under the condition of not considering disturbance, the optimization variable is historical operation sample data, and resource sample data corresponding to the optimization variable is recorded. The historical job sample data can be represented by x, the resource sample data is a function value f (x) corresponding to x, and x and f (x) are recorded by archive. In the following description, a first prediction model corresponding to a curve is a first curve, points recorded by an archive are drawn on the first curve, and a third curve formed by discrete points is obtained.
Wherein, the optimization algorithm comprises: a heritage algorithm, a differential evolution algorithm, a particle swarm optimization algorithm, a mathematical programming method and the like.
Specifically, each local optimal solution and each local worst solution in the archive are identified through a peak detection algorithm. Wherein, the local optimal solution corresponds to the wave crest, and the local worst solution corresponds to the wave trough. Wherein, the local optimal solution is marked as xpeak,i,i=1,…,npeak,npeakIs an integer, the local worst solution is denoted xvalley,j,j=1,…,nvalley,nvalleyAre integers.
Specifically, the peak and the trough may be determined on any one of the first curve and the third curve.
Specifically, the historical job sample data corresponding to each local optimal solution and each local worst solution is used as the target historical job sample data, that is, x is used as the target historical job sample datapeak,iAnd xvalley,jAs target historical job sample data.
Wherein, the formula (1) is a first preset calculation formula.
Specifically, x ispeak,iAnd xvalley,jAre respectively substituted into the formula (1) to obtain xpeak,iAnd xvalley,jRespectively corresponding robust function values, xpeak,iAnd xvalley,jThe corresponding robust function value is used as a first training parameter.
In a specific embodiment, the manner of obtaining the first target parameter is specifically as shown in fig. 4:
Wherein the first target value is used to indicate the robustness level.
Specifically, the second preset calculation formula is shown in formula (2):
wherein the content of the first and second substances,is xpeak,iThe value of the corresponding robust function is,is xvalley,jThe value of the corresponding robust function is,is a first target value.
Specifically, a first target value is input into a first prediction model, and the first target value corresponds to a first target valueTarget parameter, wherein the first target parameter isAnd (4) showing. Wherein, when the first prediction model is established, the corresponding relation between the first target value and the first target parameter is stored in advance.
And 104, constructing a second prediction model based on the target historical operation sample data and the first training parameter.
In a specific embodiment, the second prediction model is specifically constructed as shown in fig. 5:
E.g. will phixNoted as a first set:
Φx={xi,i=1,…,npeak+nvalley}
={xpeak,i,i=1,…,npeak}
∪{xvalley,j,j=1,…,nvalley}
E.g. will phifAnd as a second set:
and step 503, fitting to obtain a mapping relation between the first set and the second set.
Specifically, fitting phi using a data fitting methodxAnd phifThe mapping relationship between them.
The data fitting method comprises the following steps: a Radial Basis Function Network (RBF) algorithm, a polynomial fitting method, and the like.
And 105, determining a second target parameter corresponding to the second prediction model.
Specifically, the second prediction model is optimized by using an optimization algorithm, the optimization variable is target historical operation sample data, and target resource sample data corresponding to the optimization variable is recorded.
In a specific embodiment, a second curve corresponding to the second prediction model is drawn, and based on the second curve, target historical operation sample data corresponding to the maximum peak point is used as a second target parameter. For example, the target historical job sample data corresponding to 3.6 in fig. 2 is used as the second target parameter, wherein the second target parameter is used as the second target parameterAnd (4) showing.
And 106, adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.
In one embodiment, after obtaining the second target parameter, the second target parameter is input into equation (1) to obtain a second target value, wherein the second target value is usedAnd (4) showing.
In a specific embodiment, the implementation of the prediction model is specifically shown in fig. 6:
Specifically, the difference is an absolute value of a difference between the first target parameter and the second target parameter.
Specifically, when the difference is not within the preset range, judging whether the first target value is larger than the second target value, if so, adding the second target parameter to the first set, adding the second target value to the second set, training the second prediction model based on the added first set and the added second set until the training of the second prediction model is completed, otherwise, updating the first target value to the second target value, and updating the first target parameter to the second target parameter; and training the second prediction model based on the updated first target value and the updated first target parameter until the training of the second prediction model is completed.
According to the training method of the prediction model and the prediction method based on the prediction model, provided by the embodiment of the invention, the first prediction model is constructed and is obtained by training historical operation sample data of the operation machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target operation sample data based on the first prediction model; determining a first target parameter corresponding to the first prediction model based on the first training parameter; constructing a second prediction model based on target historical operation sample data and the first training parameter; determining a second target parameter corresponding to the second prediction model; the method comprises the steps of adjusting a first prediction model based on a first target parameter and a second target parameter to obtain a final prediction model, wherein the first prediction model is adjusted according to the first target parameter and the second target parameter, and the second prediction model is adjusted according to the first target parameter and the second target parameter to complete the training of the prediction model.
The embodiment of the present invention provides a prediction method based on a prediction model, which corresponds to the above-mentioned training method of the prediction model, and the repeated parts are not repeated, and the specific implementation is as shown in fig. 7:
in step 701, work data of the work machine during work is acquired.
Specifically, the method for acquiring the work data includes the following steps: operational data and environmental data.
Specifically, the resource data output by the prediction model is sent to a terminal installed on the excavator, so that the terminal can prompt a user in a voice or display mode, the user can operate the excavator according to the resource data, and oil consumption of the excavator is minimized.
The following describes the training device of the prediction model provided in the embodiment of the present invention, and the training device of the prediction model described below and the training method of the prediction model described above may be referred to correspondingly, and the repeated parts are not repeated, as shown in fig. 8:
a first building module 801, configured to build a first prediction model, where the first prediction model is obtained by training historical operation sample data of a work machine and resource sample data corresponding to the historical operation sample data;
a first determining module 802, configured to determine, based on the first prediction model, target historical job sample data and a first training parameter corresponding to the target historical job sample data;
a second determining module 803, configured to determine, based on the first training parameter, a first target parameter corresponding to the first prediction model;
a second construction module 804, configured to construct a second prediction model based on the target historical job sample data and the first training parameter;
a third determining module 805, configured to determine a second target parameter corresponding to the second prediction model;
an adjusting module 806 is configured to adjust the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.
In a specific embodiment, the first determining module 802 is specifically configured to determine at least one peak and at least one valley of the first prediction model; taking historical operation sample data corresponding to at least one peak and at least one trough as target historical operation sample data; and respectively inputting target historical operation sample data into a first preset calculation formula, and obtaining a first training parameter through the first preset calculation formula.
In an embodiment, the second determining module 803 is specifically configured to determine a first target value from the first training parameters based on a second preset calculation formula, where the first target value is used to indicate a level of robustness; based on the first prediction model, a first target parameter corresponding to the first target value is determined.
In a specific embodiment, the second building module 804 is specifically configured to combine sample data of the target historical job to obtain a first set; combining the first training parameters to obtain a second set; fitting to obtain a mapping relation between the first set and the second set; and constructing a second prediction model based on the target historical operation sample data, the first training parameters and the mapping relation.
In an embodiment, the third determining module 805 is specifically configured to use a position corresponding to the maximum value as the second target parameter based on the second prediction model.
In a specific embodiment, the adjusting module 806 is further configured to input the second target parameter into a first preset calculation formula, and obtain a second target value through the first preset calculation formula; an adjusting module 806, specifically configured to calculate a difference between the first target parameter and the second target parameter; judging whether the difference value is within a preset range; when the difference value is judged to be within the preset range, determining that the training of the prediction model is finished; when the difference value is not within the preset range, adding a second target parameter to the first set, adding a second target value to the second set, and training a second prediction model based on the added first set and the added second set until the training of the second prediction model is completed; when the difference is not within the preset range and the first target value is smaller than the second target value, updating the first target value to the second target value and updating the first target parameter to the second target parameter; and training the second prediction model based on the updated first target value and the updated first target parameter until the training of the second prediction model is completed.
The prediction device based on the prediction model provided in the embodiment of the present invention is described below, and the prediction device based on the prediction model described below and the prediction method based on the prediction model described above may be referred to correspondingly, and repeated parts are not repeated, as shown in fig. 9 specifically:
an obtaining module 901, configured to obtain operation data of an operation machine during operation;
an output module 902, configured to input the job data into the prediction model, obtain resource data output by the prediction model, and send the resource data to a terminal device, so that the terminal device outputs the resource data.
Fig. 10 illustrates a physical structure diagram of an electronic device, and as shown in fig. 10, the electronic device may include: a processor (processor)1001, a communication Interface (communication Interface)1002, a memory (memory)1003 and a communication bus 1004, wherein the processor 1001, the communication Interface 1002 and the memory 1003 complete communication with each other through the communication bus 1004. Processor 1001 may call logic instructions in memory 1003 to perform a method of training a predictive model, the method comprising: constructing a first prediction model, wherein the first prediction model is obtained by training historical operation sample data of the operation machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target operation sample data based on the first prediction model; determining a first target parameter corresponding to the first prediction model based on the first training parameter; constructing a second prediction model based on target historical operation sample data and the first training parameter; determining a second target parameter corresponding to the second prediction model; adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model; or, performing a prediction method based on a prediction model, the method comprising: acquiring operation data of an operation machine during operation; and inputting the operation data into the prediction model, obtaining resource data output by the prediction model, and sending the resource data to the terminal equipment so as to enable the terminal equipment to output the resource data.
In addition, the logic instructions in the memory 1003 may be implemented in the form of software functional units and may be stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute 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), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention also provides 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, enable the computer to perform a method for training a prediction model provided by the above methods, the method comprising: constructing a first prediction model, wherein the first prediction model is obtained by training historical operation sample data of the operation machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target operation sample data based on the first prediction model; determining a first target parameter corresponding to the first prediction model based on the first training parameter; constructing a second prediction model based on target historical operation sample data and the first training parameter; determining a second target parameter corresponding to the second prediction model; adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model; or, performing a prediction method based on a prediction model, the method comprising: acquiring operation data of an operation machine during operation; and inputting the operation data into the prediction model, obtaining resource data output by the prediction model, and sending the resource data to the terminal equipment so as to enable the terminal equipment to output the resource data.
In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method for performing the training of the prediction models provided above, the method comprising: constructing a first prediction model, wherein the first prediction model is obtained by training historical operation sample data of the operation machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target operation sample data based on the first prediction model; determining a first target parameter corresponding to the first prediction model based on the first training parameter; constructing a second prediction model based on target historical operation sample data and the first training parameter; determining a second target parameter corresponding to the second prediction model; adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model; or, performing a prediction method based on a prediction model, the method comprising: acquiring operation data of an operation machine during operation; and inputting the operation data into the prediction model, obtaining resource data output by the prediction model, and sending the resource data to the terminal equipment so as to enable the terminal equipment to output the resource data.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for training a predictive model, comprising:
constructing a first prediction model, wherein the first prediction model is obtained by training historical operation sample data of a working machine and resource sample data corresponding to the historical operation sample data; determining target historical operation sample data and a first training parameter corresponding to the target historical operation sample data based on the first prediction model;
determining a first target parameter corresponding to the first prediction model based on the first training parameter;
constructing a second prediction model based on the target historical operation sample data and the first training parameter;
determining a second target parameter corresponding to the second prediction model;
and adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.
2. The training method of the prediction model according to claim 1, wherein the determining, based on the first prediction model, target historical operation sample data and a first training parameter corresponding to the target historical operation sample data comprises:
determining at least one peak and at least one valley of the first predictive model;
taking historical operation sample data corresponding to the at least one peak and the at least one valley as the target historical operation sample data;
and respectively inputting the target historical operation sample data into a first preset calculation formula, and obtaining the first training parameter through the first preset calculation formula.
3. The method for training the prediction model according to claim 2, wherein the determining the first target parameter corresponding to the first prediction model based on the first training parameter comprises:
determining a first target value from the first training parameters based on a second preset calculation formula, wherein the first target value is used for indicating the robustness;
determining the first target parameter corresponding to the first target value based on a first prediction model.
4. The training method of a prediction model according to claim 3, wherein the constructing a second prediction model based on the target historical operation sample data and the first training parameters comprises:
combining the target historical operation sample data to obtain a first set;
combining the first training parameters to obtain a second set;
fitting to obtain a mapping relation between the first set and the second set;
and constructing a second prediction model based on the target historical operation sample data, the first training parameter and the mapping relation.
5. The method for training the prediction model according to claim 1, wherein the determining the second target parameter corresponding to the second prediction model comprises:
and taking the position corresponding to the maximum value as the second target parameter based on the second prediction model.
6. The method of claim 4, wherein before adjusting the second prediction model based on the first target parameter and the second target parameter to obtain the final prediction model, the method further comprises:
inputting the second target parameter into the first preset calculation formula, and obtaining a second target value through the first preset calculation formula;
adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model, including:
calculating a difference between the first target parameter and the second target parameter;
judging whether the difference value is within a preset range;
when the difference value is judged to be within the preset range, determining that the training of the prediction model is finished;
when the difference value is not determined to be within the preset range, adding the second target parameter to the first set, adding the second target value to the second set, and training the second prediction model based on the added first set and the added second set until the training of the second prediction model is completed;
when the difference is determined not to be within the preset range and the first target value is smaller than the second target value, updating the first target value to the second target value and the first target parameter to the second target parameter; training the second prediction model based on the updated first target value and the updated first target parameter until the training of the second prediction model is completed.
7. A prediction method based on the prediction model of any one of 1 to 6, comprising:
acquiring operation data of an operation machine during operation;
and inputting the operation data into the prediction model, obtaining resource data output by the prediction model, and sending the resource data to terminal equipment so as to enable the terminal equipment to output the resource data.
8. An apparatus for training a predictive model, comprising:
the system comprises a first construction module, a second construction module and a third construction module, wherein the first construction module is used for constructing a first prediction model, and the first prediction model is obtained by training historical operation sample data of a working machine and resource sample data corresponding to the historical operation sample data;
the first determination module is used for determining target historical operation sample data and a first training parameter corresponding to the target historical operation sample data based on the first prediction model;
a second determining module, configured to determine, based on the first training parameter, a first target parameter corresponding to the first prediction model;
the second construction module is used for constructing a second prediction model based on the target historical operation sample data and the first training parameter;
a third determining module, configured to determine a second target parameter corresponding to the second prediction model;
and the adjusting module is used for adjusting the second prediction model based on the first target parameter and the second target parameter so as to obtain a final prediction model.
9. A prediction apparatus based on a prediction model, comprising:
the acquisition module is used for acquiring operation data of the operation machine during operation;
and the output module is used for inputting the operation data into the prediction model, obtaining the resource data output by the prediction model, and sending the resource data to the terminal equipment so as to enable the terminal equipment to output the resource data.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program performs the steps of the method for training a prediction model according to any one of claims 1 to 6 or the steps of the method for prediction model based prediction according to claim 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111117392.6A CN113988369A (en) | 2021-09-23 | 2021-09-23 | Prediction model training method and prediction method based on prediction model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111117392.6A CN113988369A (en) | 2021-09-23 | 2021-09-23 | Prediction model training method and prediction method based on prediction model |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113988369A true CN113988369A (en) | 2022-01-28 |
Family
ID=79736425
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111117392.6A Pending CN113988369A (en) | 2021-09-23 | 2021-09-23 | Prediction model training method and prediction method based on prediction model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113988369A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114739493A (en) * | 2022-04-26 | 2022-07-12 | 上海三一重机股份有限公司 | Material weighing method and device in operation machine and operation machine |
CN115478574A (en) * | 2022-10-31 | 2022-12-16 | 吉林大学 | Excavator load prediction method based on radial basis function neural network |
-
2021
- 2021-09-23 CN CN202111117392.6A patent/CN113988369A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114739493A (en) * | 2022-04-26 | 2022-07-12 | 上海三一重机股份有限公司 | Material weighing method and device in operation machine and operation machine |
CN114739493B (en) * | 2022-04-26 | 2024-01-26 | 上海三一重机股份有限公司 | Material weighing method and device in working machine and working machine |
CN115478574A (en) * | 2022-10-31 | 2022-12-16 | 吉林大学 | Excavator load prediction method based on radial basis function neural network |
CN115478574B (en) * | 2022-10-31 | 2024-03-19 | 吉林大学 | Excavator load prediction method based on radial basis function neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113988369A (en) | Prediction model training method and prediction method based on prediction model | |
CN109379240B (en) | Internet of vehicles flow prediction model construction method and device and electronic equipment | |
CN105678395B (en) | Neural network establishing method and system and neural network application method and system | |
CN112861362A (en) | Power assembly performance parameter optimization method and device based on vehicle oil consumption | |
CN113705929A (en) | Spring festival holiday load prediction method based on load characteristic curve and typical characteristic value fusion | |
CN113236431B (en) | Engineering machinery energy-saving control method and device and engineering machinery | |
CN111122222A (en) | Sample point position determining method and system | |
US20190392312A1 (en) | Method for quantizing a histogram of an image, method for training a neural network and neural network training system | |
CN113641525A (en) | Variable exception recovery method, apparatus, medium, and computer program product | |
CN113610709B (en) | Model quantization method, apparatus, electronic device, and computer-readable storage medium | |
CN117035482A (en) | Land utilization comprehensive renovation performance evaluation method and system | |
CN109416748B (en) | SVM-based sample data updating method, classification system and storage device | |
CN114753940A (en) | Engine speed control method, electronic device, engineering machine and storage medium | |
CN111767981B (en) | Approximate calculation method of Mish activation function | |
CN110417020B (en) | Load flow calculation method and system for comprehensive energy system for processing non-smooth constraint | |
CN114004873A (en) | Method, device, equipment, medium and product for detecting flatness of operation area | |
CN112116076A (en) | Optimization method and optimization device for activation function | |
CN112633617A (en) | Maintenance strategy generation method and device, computer equipment and readable storage medium | |
CN113036769A (en) | Static voltage stability analysis method and system for power system | |
CN113774985B (en) | Method and device for controlling working machine, medium, and working machine | |
CN111552197A (en) | BIM-based refrigeration equipment energy consumption optimization method, device and equipment | |
CN112200730A (en) | Image filtering processing method, device, equipment and storage medium | |
CN117726807B (en) | Infrared small target detection method and system based on scale and position sensitivity | |
CN117313438B (en) | Three-dimensional factory tool editing optimization method and device | |
CN113978663B (en) | Automatic ballast water correction system and method and readable 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 |