Training method of prediction model and prediction method based on prediction model

Abstract

The embodiment of the present invention provides a prediction model training method and a prediction method based on the prediction model, the training method includes: constructing a first prediction model, the first prediction model is obtained by training the historical operation sample data of the operation machinery and the resource sample data corresponding to the historical operation sample data; based on the first prediction model, determining the target historical operation sample data and the first training parameter corresponding to the target operation sample data; based on the first training parameter, determining the first target parameter corresponding to the first prediction model; based on the target historical operation sample data and the first training parameter, constructing a second prediction model; determining the second target parameter corresponding to the second prediction model; based on the first target parameter and the second target parameter, adjusting the second prediction model to obtain the final prediction model. The present invention is used to solve the defects of low efficiency and high energy consumption of excavators in the prior art when operating complex and changeable operation objects.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a prediction model training method and a prediction method based on the prediction model. Background Art

[0002] As a kind of engineering machinery with high load-bearing capacity and operating on complex ground conditions, the operating environment and operating objects of excavators are complex and changeable, the power output intensity is large, and the fuel consumption is very large. Therefore, it is very important to optimize the operating state of the excavator according to the operating requirements of the excavator to improve the operating efficiency of the excavator and reduce the fuel consumption of the excavator. In addition, due to the complexity and changeability of the operating objects, such as the uneven flatness of the operating object leading to changes in the travel speed, the different hardness of the operating object leading to changes in the excavation speed, etc., the operating state of the excavator cannot be maintained under a certain constant condition.

[0003] Therefore, when facing complex and changeable working objects, how to provide the excavator with corresponding operating strategies in a timely manner, so that the excavator can adapt to the current working objects, and at the same time improve working efficiency and reduce fuel consumption while adapting to the working objects is an important issue that needs to be urgently solved in the industry. Summary of the invention

[0004] The embodiments of the present invention provide a prediction model training method and a prediction method based on the prediction model, so as to solve the defects of low efficiency and high energy consumption of excavators when operating on complex and changeable working objects in the prior art, so as to enable the excavators to improve working efficiency and reduce fuel consumption while adapting to the working objects.

[0005] An embodiment of the present invention provides a prediction model training method, comprising:

[0006] Constructing a first prediction model, wherein the first prediction model is obtained by training with 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 first training parameters corresponding to the target historical operation sample data based on the first prediction model;

[0007] Determining a first target parameter corresponding to the first prediction model based on the first training parameter;

[0008] Building a second prediction model based on the target historical operation sample data and the first training parameters;

[0009] Determining a second target parameter corresponding to the second prediction model;

[0010] Based on the first target parameter and the second target parameter, the second prediction model is adjusted to obtain a final prediction model.

[0011] According to a prediction model training method according to an embodiment of the present invention, determining target historical job sample data and first training parameters corresponding to the target historical job sample data based on the first prediction model includes:

[0012] determining at least one peak and at least one trough of the first prediction model;

[0013] Using the historical operation sample data corresponding to the at least one peak and the at least one trough as the target historical operation sample data;

[0014] The target historical operation sample data are respectively input into a first preset calculation formula, and the first training parameter is obtained through the first preset calculation formula.

[0015] According to a prediction model training method according to an embodiment of the present invention, determining a first target parameter corresponding to the first prediction model based on the first training parameter includes:

[0016] Based on a second preset calculation formula, determining a first target value from the first training parameter, the first target value being used to indicate the level of robustness;

[0017] Based on the first prediction model, the first target parameter corresponding to the first target value is determined.

[0018] According to a prediction model training method according to an embodiment of the present invention, constructing a second prediction model based on the target historical job sample data and the first training parameters includes:

[0019] Combining the target historical operation sample data to obtain a first set;

[0020] combining the first training parameters to obtain a second set;

[0021] Fitting to obtain a mapping relationship between the first set and the second set;

[0022] A second prediction model is constructed based on the target historical job sample data, the first training parameters and the mapping relationship.

[0023] According to the prediction model training method of one embodiment of the present invention, the determining the second target parameter corresponding to the second prediction model includes:

[0024] Based on the second prediction model, the position corresponding to the maximum value is used as the second target parameter.

[0025] According to the prediction model training method of one embodiment of the present invention, before adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model, the method further includes:

[0026] Inputting the second target parameter into the first preset calculation formula, and obtaining a second target value through the first preset calculation formula;

[0027] The adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model includes:

[0028] calculating a difference between the first target parameter and the second target parameter;

[0029] Determining whether the difference is within a preset range;

[0030] When it is determined that the difference is within the preset range, determining that the prediction model training is completed;

[0031] When it is determined that the difference is not within the preset range, the second target parameter is added to the first set, the second target value is added to the second set, and the second prediction model is trained based on the added first set and the added second set until the training of the second prediction model is completed;

[0032] When it is determined that the difference is not within the preset range and the first target value is less than the second target value, the first target value is updated to the second target value, and the first target parameter is updated to the second target parameter; based on the updated first target value and the updated first target parameter, the second prediction model is trained until the training of the second prediction model is completed.

[0033] The embodiment of the present invention further provides a prediction method based on the prediction model, comprising:

[0034] Obtaining the operation data of the operating machine during operation;

[0035] The operation data is input into the prediction model, resource data output by the prediction model is obtained, and the resource data is sent to a terminal device so that the terminal device outputs the resource data.

[0036] The embodiment of the present invention also provides a prediction model training device, including:

[0037] A first construction module is used to construct a first prediction model, wherein the first prediction model is obtained by training the historical operation sample data of the operation machine and the resource sample data corresponding to the historical operation sample data;

[0038] A first determination module, configured to determine target historical operation sample data and a first training parameter corresponding to the target historical operation sample data based on the first prediction model;

[0039] A second determination module, configured to determine a first target parameter corresponding to the first prediction model based on the first training parameter;

[0040] A second construction module, configured to construct a second prediction model based on the target historical job sample data and the first training parameters;

[0041] A third determination module, used to determine a second target parameter corresponding to the second prediction model;

[0042] An adjustment module is used to adjust the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.

[0043] The embodiment of the present invention further provides a prediction device based on a prediction model, comprising:

[0044] An acquisition module is used to acquire the operation data of the operation machine during operation;

[0045] The output module is used to input the operation data into the prediction model, obtain the resource data output by the prediction model, and send the resource data to the terminal device so that the terminal device outputs the resource data.

[0046] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of a training method for a prediction model as described above or the steps of a prediction method based on a prediction model are implemented.

[0047] The training method of the prediction model and the prediction method based on the prediction model provided by the embodiment of the present invention are as follows: by constructing a first prediction model, the first prediction model is obtained by training the historical operation sample data of the operating machinery and the resource sample data corresponding to the historical operation sample data; based on the first prediction model, the target historical operation sample data and the first training parameter corresponding to the target operation sample data are determined; based on the first training parameter, the first target parameter corresponding to the first prediction model is determined; based on the target historical operation sample data and the first training parameter, a second prediction model is constructed; the second target parameter corresponding to the second prediction model is determined; based on the first target parameter and the second target parameter, the second prediction model is adjusted to obtain the final prediction model. It can be seen that the present invention avoids a large amount of sampling in the training process of the second prediction model, and uses the first target parameter and the second target parameter to adjust the second prediction model to complete the training of the prediction model, thereby improving the model training efficiency. In addition, the present invention inputs the operation data of the operating machinery into the constructed prediction model to output resource data to provide the operator with a solution on how to operate the industrial machinery. The prediction model constructed by the present invention can output the corresponding resource data in a timely manner for various complex and changeable operating objects, so that the operating machinery can adapt to various operating objects, and while adapting to the operating objects, the working efficiency of the operating machinery can be improved and the fuel consumption of the operating machinery can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 It is one of the flowcharts of the training method of the prediction model provided by the embodiment of the present invention;

[0050] Figure 2 This is a second flow chart of the training method of the prediction model provided by the embodiment of the present invention;

[0051] Figure 3 This is a third flow chart of the training method of the prediction model provided by the embodiment of the present invention;

[0052] Figure 4 This is a fourth flow chart of the training method of the prediction model provided by the embodiment of the present invention;

[0053] Figure 5 This is a fifth flow chart of the training method of the prediction model provided by the embodiment of the present invention;

[0054] Figure 6This is a sixth flow chart of the training method of the prediction model provided by the embodiment of the present invention;

[0055] Figure 7 is a flowchart of a prediction method provided by an embodiment of the present invention;

[0056] Figure 8 It is a schematic diagram of the structure of a training device for a prediction model provided by an embodiment of the present invention;

[0057] Fig. 9 is a schematic diagram of the structure of a prediction device provided by an embodiment of the present invention;

[0058] Fig.10 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] Combine the following Figures 1 to 6 The following describes a method for training a prediction model according to an embodiment of the present invention.

[0061] The embodiment of the present invention provides a method for training a prediction model, which can be applied to intelligent terminals, such as mobile phones, computers, tablets, etc., and can also be applied to servers, and can also be applied to controllers of operating machinery. Below, the method is applied to a server as an example for explanation, but it should be noted that it is only an example and is not used to limit the scope of protection of the present invention. Some other descriptions in the embodiments of the present invention are also examples and are not used to limit the scope of protection of the present invention, and will not be described one by one later.

[0062] The specific implementation of the training method of the prediction model is as follows Figure 1 As shown:

[0063] Step 101: construct a first prediction model.

[0064] The first prediction model is obtained by training the historical operation sample data of the operating machinery and the resource sample data corresponding to the historical operation sample data.

[0065] Among them, operating machinery includes: excavators, electric excavators, hydraulic excavators, loaders, etc.

[0066] Specifically, the historical operation sample data includes: the operation data of the excavator obtained by the sensor installed on the excavator, and the environmental data of the working object when the excavator is working. The operation data includes: the excavation speed of the excavator, the transfer speed of the excavator, etc., and the environmental data includes: the flatness of the ground of the working object, the hardness of the ground of the working object, etc.

[0067] Specifically, the resource sample data includes: fuel consumption of the excavator, power consumption of the excavator, time consumption of the excavator, etc.

[0068] Below, an example is given of an operation scene where an excavator is leveling the ground, the historical operation sample data is the transfer speed of the excavator, and the resource sample data is the fuel consumption of the excavator. It should be noted that this embodiment is only an example and is not intended to limit the scope of protection of the present invention.

[0069] Specifically, before the detailed description is given, the present invention will first explain the relevant definitions.

[0070] First, a first objective function and a second objective function are defined, wherein the first objective function corresponds to the first prediction model, and the second objective function corresponds to the second prediction model. Figure 2 The function corresponding to the thin curve (first curve) in is the first objective function f(x), Figure 2 The function corresponding to the medium-thick curve (second curve) is the second objective function f eff (x) It should be noted that Figure 2 The thick and thin curves in are only used to distinguish the two functions and have no other meanings.

[0071] The present invention is to minimize the fuel consumption of the excavator by finding the point marked by the five-pointed star in the second curve. Below, the robust solution is defined: corresponding to the maximization of the objective function f(x), a solution x * The prerequisite for being called a robust solution is that x * It is the global optimal solution of formula (1).

[0072]

[0073] in,

[0074] Where x = [x 1 ,x 2 ,…,x D ] T is the decision variable, i.e., the transfer speed of the excavator, Ω is the feasible domain of the decision variable, and f(x,δ) is the robust function value. f(x,δ) can also be expressed as f eff (x) indicates that is the δ neighborhood of x, δ = [δ1 ,δ 2 ,…,δ D ] T represents the disturbance intensity of the decision variable, H is the number of samples set by the user, and k is the variable.

[0075] in, Figure 2 An example is given for explanation when δ=1.

[0076] For example, when δ=1 and the speed is 3.6, the specific calculation steps of the robust function value are:

[0077] Determine the fuel consumption corresponding to the preset number of speeds in the range [2.6-4.6]. For example, the preset number is 10000. Figure 2 The dots in the dotted line area of ​​. Calculate the average value of 10,000 fuel consumption values ​​and use the average value as the robust function value.

[0078] Step 102: Determine target historical operation sample data and first training parameters corresponding to the target historical operation sample data based on the first prediction model.

[0079] In a specific embodiment, the target historical operation sample data and the first training parameter are obtained as follows: Figure 3 As shown:

[0080] Step 301: determine at least one peak and at least one trough of a first prediction model.

[0081] Specifically, the first prediction model is optimized by using an optimization algorithm without considering disturbances, the optimization variable is the historical operation sample data, and the resource sample data corresponding to the optimization variable is recorded. The historical operation sample data can be represented by x, and the resource sample data is the function value f(x) corresponding to x, and x and f(x) are recorded by archive. The first prediction model is taken as a smooth curve as an example for explanation, and the curve corresponding to the first prediction model is the first curve. On the basis of the first curve, the points recorded in the archive are depicted to obtain a third curve composed of discrete points.

[0082] Among them, the optimization algorithms include: heritage algorithm, differential evolution algorithm, particle swarm optimization algorithm and mathematical programming method.

[0083] Specifically, the peak detection algorithm is used to identify each local optimal solution and each local worst solution in the archive. The local optimal solution corresponds to the peak, and the local worst solution corresponds to the trough. The local optimal solution is denoted by x peak,i , i=1,…,n peak , n peak is an integer, and the local worst solution is denoted by xvalley,j , j=1,…,n valley , n valley is an integer.

[0084] Specifically, the peaks and troughs may be determined on either the first curve or the third curve.

[0085] Step 302: Use historical operation sample data corresponding to at least one peak and at least one trough as target historical operation sample data.

[0086] Specifically, the historical operation sample data corresponding to each local optimal solution and each local worst solution is used as the target historical operation sample data, that is, x peak,i and x valley,j As the target historical job sample data.

[0087] Step 303: Input the target historical operation sample data into a first preset calculation formula respectively, and obtain a first training parameter through the first preset calculation formula.

[0088] Wherein, formula (1) is the first preset calculation formula.

[0089] Specifically, x peak,i and x valley,j Substitute into formula (1) to get x peak,i and x valley,j The corresponding robust function values ​​are x peak,i and x valley,j The corresponding robust function value is used as the first training parameter.

[0090] Step 103: Determine a first target parameter corresponding to the first prediction model based on the first training parameter.

[0091] In a specific embodiment, the first target parameter is obtained as follows: Figure 4 As shown:

[0092] Step 401, determining a first target value from a first training parameter based on a second preset calculation formula.

[0093] The first target value is used to indicate the level of robustness.

[0094] Specifically, the second preset calculation formula is shown in formula (2):

[0095]

[0096] in, For x peak,i The corresponding robust function value is, For x valley,j The corresponding robust function value is, is the first target value.

[0097] Step 402: Determine a first target parameter corresponding to a first target value based on a first prediction model.

[0098] Specifically, the first target value is input into the first prediction model, and the first target parameter corresponding to the first target value is expressed as In other words, when the first prediction model is established, the corresponding relationship between the first target value and the first target parameter has been pre-stored.

[0099] Step 104: construct a second prediction model based on the target historical operation sample data and the first training parameters.

[0100] In a specific embodiment, the second prediction model is constructed as follows: Figure 5 As shown:

[0101] Step 501: Combine target historical job sample data to obtain a first set.

[0102] For example, Φ x Let’s call it the first set:

[0103] Φ x ={x i ,i=1,…,n peak +n valley}

[0104] ={x peak,i ,i=1,…,n peak}

[0105] ∪{x valley,j ,j=1,…,n valley}

[0106] Step 502: Combine the first training parameters to obtain a second set.

[0107] For example, Φ f Let’s call it the second set:

[0108]

[0109] Step 503: fitting to obtain a mapping relationship between the first set and the second set.

[0110] Specifically, using the data fitting method, fitting Φ x and Φ f The mapping relationship between them.

[0111] Among them, the data fitting methods include: Radial Basis Function Network algorithm (Radial Basis Function Network, referred to as RBF) and polynomial fitting method.

[0112] Step 504: construct a second prediction model based on the target historical job sample data, the first training parameter and the mapping relationship.

[0113] Step 105: Determine a second target parameter corresponding to the second prediction model.

[0114] Specifically, the second prediction model is optimized using an optimization algorithm, the optimization variables are the target historical job sample data, and the target resource sample data corresponding to the optimization variables are recorded.

[0115] In a specific embodiment, a second curve corresponding to the second prediction model is drawn, and based on the second curve, the target historical operation sample data corresponding to the maximum peak point is used as the second target parameter. Figure 2 The target historical operation sample data corresponding to 3.6 is used as the second target parameter, where the second target parameter is express.

[0116] Step 106: Adjust the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.

[0117] In a specific embodiment, after obtaining the second target parameter, the second target parameter is input into formula (1) to obtain the second target value, wherein the second target value is expressed as express.

[0118] In a specific embodiment, the prediction model is implemented as follows: Figure 6 As shown:

[0119] Step 601, calculating the difference between the first target parameter and the second target parameter.

[0120] Specifically, the difference is the absolute value of the difference between the first target parameter and the second target parameter.

[0121] Step 602, determine whether the difference is within a preset range, if so, execute step 603, otherwise, execute steps 604 and 605.

[0122] Step 603, determine whether the prediction model training is completed, and output the first target parameter and the first target value.

[0123] Step 604, add the second target parameter to the first set, add the second target value to the second set, and train 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.

[0124] Step 605, when it is determined that the difference is not within the preset range and the first target value is less than the second target value, the first target value is updated to the second target value, and the first target parameter is updated to the second target parameter; based on the updated first target value and the updated first target parameter, the second prediction model is trained until the training of the second prediction model is completed.

[0125] Specifically, when the difference is not within the preset range, determine whether the first target value is greater than the second target value. If so, add the second target parameter to the first set, add the second target value to the second set, and train 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, update the first target value to the second target value, and update the first target parameter to the second target parameter; train 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.

[0126] The training method of the prediction model and the prediction method based on the prediction model provided by the embodiment of the present invention are as follows: by constructing a first prediction model, the first prediction model is obtained by training the historical operation sample data of the operating machinery and the resource sample data corresponding to the historical operation sample data; based on the first prediction model, the target historical operation sample data and the first training parameter corresponding to the target operation sample data are determined; based on the first training parameter, the first target parameter corresponding to the first prediction model is determined; based on the target historical operation sample data and the first training parameter, a second prediction model is constructed; the second target parameter corresponding to the second prediction model is determined; based on the first target parameter and the second target parameter, the second prediction model is adjusted to obtain the final prediction model. It can be seen that the present invention avoids a large amount of sampling in the training process of the second prediction model, and uses the first target parameter and the second target parameter to adjust the second prediction model to complete the training of the prediction model, thereby improving the model training efficiency. In addition, the present invention inputs the operation data of the operating machinery into the constructed prediction model to output resource data to provide the operator with a solution on how to operate the industrial machinery. The prediction model constructed by the present invention can output the corresponding resource data in a timely manner for various complex and changeable operating objects, so that the operating machinery can adapt to various operating objects, and while adapting to the operating objects, the working efficiency of the operating machinery can be improved and the fuel consumption of the operating machinery can be reduced.

[0127] The embodiment of the present invention provides a prediction method based on a prediction model, which corresponds to the training method of the prediction model described above, and the repetitive parts are not repeated here. The specific implementation is as follows: Figure 7 As shown:

[0128] Step 701, obtaining the operation data of the operation machine during operation.

[0129] Specifically, the operating data obtained by the sensor installed on the excavator is obtained, and the operating data includes: operation data and environmental data.

[0130] Step 702, 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 device so that the terminal device outputs the resource data.

[0131] Specifically, the resource data output by the prediction model is sent to a terminal installed on the excavator, so that the terminal can prompt the user through voice or display, so that the user can operate the excavator according to the resource data to minimize the fuel consumption of the excavator.

[0132] The following is a description of a training device for a prediction model provided by an embodiment of the present invention. The training device for a prediction model described below and the training method for a prediction model described above can be referred to each other, and the repeated parts will not be repeated. Figure 8 As shown:

[0133] The first construction module 801 is used to construct a first prediction model, which is obtained by training the historical operation sample data of the operation machine and the resource sample data corresponding to the historical operation sample data;

[0134] A first determination module 802 is used to determine target historical operation sample data and first training parameters corresponding to the target historical operation sample data based on a first prediction model;

[0135] A second determination module 803 is used to determine a first target parameter corresponding to the first prediction model based on the first training parameter;

[0136] A second construction module 804 is used to construct a second prediction model based on the target historical job sample data and the first training parameter;

[0137] The third determination module 805 is used to determine the second target parameter corresponding to the second prediction model;

[0138] The adjustment module 806 is used to adjust the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model.

[0139] In a specific embodiment, the first determination module 802 is specifically used to determine at least one peak and at least one trough of the first prediction model; use the historical operation sample data corresponding to the at least one peak and at least one trough as the target historical operation sample data; input the target historical operation sample data into the first preset calculation formula respectively, and obtain the first training parameter through the first preset calculation formula.

[0140] In a specific embodiment, the second determination module 803 is specifically used to determine a first target value from the first training parameter based on a second preset calculation formula, where the first target value is used to indicate the level of robustness; and to determine a first target parameter corresponding to the first target value based on the first prediction model.

[0141] In a specific embodiment, the second construction module 804 is specifically used to combine the target historical job sample data to obtain a first set; combine the first training parameters to obtain a second set; fit the mapping relationship between the first set and the second set; and construct a second prediction model based on the target historical job sample data, the first training parameters and the mapping relationship.

[0142] In a specific embodiment, the third determination module 805 is specifically configured to use the position corresponding to the maximum value as the second target parameter based on the second prediction model.

[0143] In a specific embodiment, the adjustment module 806 is also used to input the second target parameter into the first preset calculation formula, and obtain the second target value through the first preset calculation formula; the adjustment module 806 is specifically used to calculate the difference between the first target parameter and the second target parameter; determine whether the difference is within a preset range; when it is determined that the difference is within the preset range, determine that the prediction model training is completed; when it is determined that the difference is not within the preset range, add the second target parameter to the first set, add the second target value to the second set, and train the second prediction model based on the added first set and the added second set until the second prediction model training is completed; when it is determined that the difference is not within the preset range, and the first target value is less than the second target value, update the first target value to the second target value, and update the first target parameter to the second target parameter; based on the updated first target value and the updated first target parameter, train the second prediction model until the second prediction model training is completed.

[0144] The following is a description of a prediction device based on a prediction model provided by an embodiment of the present invention. The prediction device based on a prediction model described below and the prediction method based on a prediction model described above can be referred to each other, and the repeated parts will not be repeated. Fig. 9 As shown:

[0145] The acquisition module 901 is used to acquire the operation data of the operation machine during operation;

[0146] The output module 902 is used to input the operation data into the prediction model, obtain the resource data output by the prediction model, and send the resource data to the terminal device so that the terminal device outputs the resource data.

[0147] Fig.10 An example of a physical structure diagram of an electronic device is shown in FIG. Fig.10As shown, the electronic device may include: a processor (processor) 1001, a communication interface (Communications Interface) 1002, a memory (memory) 1003 and a communication bus 1004, wherein the processor 1001, the communication interface 1002, and the memory 1003 communicate with each other through the communication bus 1004. Processor 1001 can call the logic instructions in memory 1003 to execute a training method for a prediction model, which method includes: constructing a first prediction model, which is trained by historical operation sample data of an operating machine and resource sample data corresponding to the historical operation sample data; based on the first prediction model, determining the target historical operation sample data and the first training parameter corresponding to the target operation sample data; based on the first training parameter, determining the first target parameter corresponding to the first prediction model; based on the target historical operation sample data and the first training parameter, constructing a second prediction model; determining the 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, executing a prediction method based on a prediction model, which method includes: obtaining the operation data of the operating machine during operation; 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 device so that the terminal device outputs the resource data.

[0148] In addition, the logic instructions in the above-mentioned memory 1003 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0149] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the prediction model training method provided by the above-mentioned methods, and the method includes: constructing a first prediction model, and the first prediction model is trained by the historical operation sample data of the operating machinery and the resource sample data corresponding to the historical operation sample data; based on the first prediction model, determining the target historical operation sample data and the first training parameter corresponding to the target operation sample data; based on the first training parameter, determining the first target parameter corresponding to the first prediction model; based on the target historical operation sample data and the first training parameter, constructing a second prediction model; determining the 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, executing a prediction method based on a prediction model, the method includes: obtaining the operation data of the operating machinery during operation; 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 device so that the terminal device outputs the resource data.

[0150] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the training methods of the above-mentioned prediction models, the method comprising: constructing a first prediction model, the first prediction model being trained by historical operation sample data of an operating machine and resource sample data corresponding to the historical operation sample data; based on the first prediction model, determining target historical operation sample data and a first training parameter corresponding to the target operation sample data; based on the first training parameter, determining a first target parameter corresponding to the first prediction model; based on the target historical operation sample data and the first training parameter, constructing a second prediction model; 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, executing a prediction method based on a prediction model, the method comprising: obtaining operation data of the operating machine during operation; inputting the operation data into the prediction model, obtaining resource data output by the prediction model, and sending the resource data to a terminal device so that the terminal device outputs the resource data.

[0151] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0152] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A prediction model training method, characterized in that: include: Constructing a first prediction model, wherein the first prediction model is obtained by training the historical operation sample data of the operation machine and the resource sample data corresponding to the historical operation sample data; Determine target historical operation sample data and first training parameters corresponding to the target historical operation sample data based on the first prediction model; The historical operation sample data includes: the operation data of the operation machine and the environmental data of the operation object when the operation machine is operating; the operation data includes: the excavation speed of the operation machine and the transfer speed of the operation machine; the environmental data includes: the ground flatness of the operation object and the ground hardness of the operation object; the resource sample data includes: the fuel consumption of the operation machine, the power consumption of the operation machine, and the time consumption of the operation machine; Determine a first target parameter corresponding to the first prediction model based on the first training parameter, wherein the first training parameter is a robust function value obtained based on the target historical operation sample data, and the first target parameter is a target robust function value obtained based on the robust function value obtained based on the target historical operation sample data; Building a second prediction model based on the target historical operation sample data and the first training parameters; Determine a second target parameter corresponding to the second prediction model; the second target parameter is the target historical operation sample data corresponding to the maximum peak point in the second curve corresponding to the second prediction model; Based on the first target parameter and the second target parameter, adjusting the second prediction model to obtain a final prediction model, wherein the prediction model is used to predict resource data corresponding to the operation object; Before adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model, the method further includes: inputting the second target parameter into the first preset calculation formula to obtain a second target value through the first preset calculation formula; The adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model includes: calculating a difference between the first target parameter and the second target parameter; Determining whether the difference is within a preset range; When it is determined that the difference is within the preset range, determining that the prediction model training is completed; When it is determined that the difference is not within the preset range, the second target parameter is added to the first set, the second target value is added to the second set, and the second prediction model is trained based on the added first set and the added second set until the training of the second prediction model is completed; wherein the first set is a set obtained by combining the target historical job sample data, and the second set is a set obtained by combining the first training parameters; When it is determined that the difference is not within the preset range and the first target value is less than the second target value, the first target value is updated to the second target value, and the first target parameter is updated to the second target parameter; based on the updated first target value and the updated first target parameter, the second prediction model is trained until the training of the second prediction model is completed; The determining of target historical operation sample data and first training parameters corresponding to the target historical operation sample data based on the first prediction model includes: determining at least one peak and at least one trough of the first prediction model; Using the historical operation sample data corresponding to the at least one peak and the at least one trough as the target historical operation sample data; Inputting the target historical operation sample data into a first preset calculation formula respectively, and obtaining the first training parameter through the first preset calculation formula; The first preset calculation formula is shown in formula (1): Formula (1) in, , is the decision variable, i.e., the transfer speed of the excavator, is the feasible domain of decision variables, is the robust function value, yes of Neighborhood, represents the disturbance intensity of the decision variable, H is the number of samples set by the user, is a variable.

2. The prediction model training method according to claim 1, characterized in that: The determining, based on the first training parameter, a first target parameter corresponding to the first prediction model includes: Based on a second preset calculation formula, determining a first target value from the first training parameter, the first target value being used to indicate the level of robustness; Determining the first target parameter corresponding to the first target value based on the first prediction model; The second preset calculation formula is shown in formula (2): Formula (2) in, is the local optimal solution for the resource sample data The corresponding robust function value is, is the local worst solution of the resource sample data The corresponding robust function value is, is the first target value.

3. The prediction model training method according to claim 2, characterized in that: The constructing a second prediction model based on the target historical job 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 relationship between the first set and the second set; A second prediction model is constructed based on the target historical job sample data, the first training parameters and the mapping relationship.

4. The prediction model training method according to claim 1, characterized in that: The determining a second target parameter corresponding to the second prediction model includes: Based on the second prediction model, the position corresponding to the maximum value is used as the second target parameter.

5. A prediction method based on the prediction model described in any one of 1-4, characterized in that: include: Obtaining the operation data of the operating machine during operation; The operation data is input into the prediction model, resource data output by the prediction model is obtained, and the resource data is sent to a terminal device so that the terminal device outputs the resource data.

6. A training device for a prediction model, characterized in that: include: A first construction module is used to construct a first prediction model, wherein the first prediction model is obtained by training the historical operation sample data of the operation machine and the resource sample data corresponding to the historical operation sample data; The historical operation sample data includes: the operation data of the operation machine and the environmental data of the operation object when the operation machine is operating; the operation data includes: the excavation speed of the operation machine and the transfer speed of the operation machine; the environmental data includes: the ground flatness of the operation object and the ground hardness of the operation object; the resource sample data includes: the fuel consumption of the operation machine, the power consumption of the operation machine, and the time consumption of the operation machine; A first determination module, configured to determine 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 determination module is used to determine a first target parameter corresponding to the first prediction model based on the first training parameter, the first training parameter being a robust function value obtained based on the target historical operation sample data, and the first target parameter being a target robust function value obtained based on the robust function value obtained based on the target historical operation sample data; A second construction module, configured to construct a second prediction model based on the target historical job sample data and the first training parameters; A third determination module is used to determine a second target parameter corresponding to the second prediction model; the second target parameter is the target historical operation sample data corresponding to the maximum peak point in the second curve corresponding to the second prediction model; an adjustment module, configured to adjust the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model, wherein the prediction model is used to predict resource data corresponding to the operation object; Before adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model, the method further includes: inputting the second target parameter into the first preset calculation formula to obtain a second target value through the first preset calculation formula; The adjusting the second prediction model based on the first target parameter and the second target parameter to obtain a final prediction model includes: calculating a difference between the first target parameter and the second target parameter; Determining whether the difference is within a preset range; When it is determined that the difference is within the preset range, determining that the prediction model training is completed; When it is determined that the difference is not within the preset range, the second target parameter is added to the first set, the second target value is added to the second set, and the second prediction model is trained based on the added first set and the added second set until the training of the second prediction model is completed; wherein the first set is a set obtained by combining the target historical job sample data, and the second set is a set obtained by combining the first training parameters; When it is determined that the difference is not within the preset range and the first target value is less than the second target value, the first target value is updated to the second target value, and the first target parameter is updated to the second target parameter; based on the updated first target value and the updated first target parameter, the second prediction model is trained until the training of the second prediction model is completed; The determining of target historical operation sample data and first training parameters corresponding to the target historical operation sample data based on the first prediction model includes: determining at least one peak and at least one trough of the first prediction model; Using the historical operation sample data corresponding to the at least one peak and the at least one trough as the target historical operation sample data; Inputting the target historical operation sample data into a first preset calculation formula respectively, and obtaining the first training parameter through the first preset calculation formula; The first preset calculation formula is shown in formula (1): Formula (1) in, , is the decision variable, i.e., the transfer speed of the excavator, is the feasible domain of decision variables, is the robust function value, yes of Neighborhood, represents the disturbance intensity of the decision variable, H is the number of samples set by the user, is a variable.

7. A prediction device based on a prediction model, characterized in that: include: An acquisition module is used to acquire the operation data of the operation machine during operation; An output module is used to input the operation data into the prediction model, obtain the 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. The prediction model is trained based on the method described in any one of claims 1 to 4.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the steps of the training method of the prediction model as described in any one of claims 1 to 4, or implements the steps of the prediction method based on the prediction model as described in claim 5.

Images