A machine learning-based intelligent control system and method for rotary tiller parameters

Abstract

The present invention relates to the field of intelligent control technology, and specifically to a rotary tiller parameter intelligent control system and method based on machine learning. The system includes a rotary tiller operating parameter data acquisition module, a rotary tiller operating parameter data processing module, a final random forest construction module, a rotary tiller abnormal operating parameter data identification module, a rotary tiller abnormal operating state determination module and a rotary tiller operating parameter tuning module; by processing the collected real-time operating parameter data of the rotary tiller, standard real-time operating parameter data of the rotary tiller is obtained, and a random forest model is constructed to identify and classify the standard real-time operating parameter data of the rotary tiller, to generate abnormal operating parameter data of the rotary tiller, and based on the abnormal operating parameter data of the rotary tiller, it is determined whether the rotary tiller is in an abnormal operating state; when the rotary tiller is in an abnormal operating state, the multiverse algorithm is used to adjust the operating parameters of the rotary tiller to achieve intelligent control of the parameters of the rotary tiller.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to a system and method for intelligently controlling parameters of a rotary tiller based on machine learning. Background Art

[0002] A rotary tiller is a land preparation machine with a wide range of applications. Compared with traditional tools, a rotary tiller can complete traditional plowing, harrowing, hoeing and leveling operations in one operation, reducing the number of times the tractor enters the land, reducing the squeeze on the land, and improving work efficiency. However, the operating parameters of the rotary tiller during operation are often adjusted manually, and the operating status of the rotary tiller cannot be adjusted intelligently, making it difficult to ensure the accuracy of the adjustment results.

[0003] The Chinese invention patent application with publication number CN117148741A introduces a method and system for intelligent control of bearing processing parameters. The method interacts with the information of the bearing to be processed and the processing equipment, inputs the digital model of the bearing cutting processing to configure the cutting control parameters and mapping auxiliary parameters, performs bearing processing control and receives feedback monitoring information, determines the data offset result in combination with the mapping auxiliary parameters and inputs the initialized optimization model, outputs the optimization control result for processing control, improves the processing control accuracy and ensures the bearing processing quality. However, by analyzing whether the feedback monitoring information is offset through the set mapping auxiliary parameters, and judging whether to perform parameter control based on the analysis results, the accuracy of the analysis results cannot be guaranteed, and it is difficult to achieve intelligent parameter control. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In order to solve the deficiencies in the background technology, the present invention provides a system and method for intelligent control of rotary tiller parameters based on machine learning, which improves the speed and accuracy of determining abnormal states of rotary tillers and realizes intelligent control of rotary tiller parameters.

[0006] (II) Technical solution

[0007] A method for intelligently controlling parameters of a rotary tiller based on machine learning comprises the following steps:

[0008] S1, setting a parameter data collection time point set and a rotary tiller operation parameter category set, and collecting the operation parameter data of the rotary tiller in real time during the operation process according to the parameter data collection time point set and the rotary tiller operation parameter category set to obtain a real-time operation parameter data matrix of the rotary tiller;

[0009] S2, performing numerical processing on the real-time operating parameter data of the rotary tiller in the real-time operating parameter data matrix of the rotary tiller to generate standard real-time operating parameter data of the rotary tiller;

[0010] S3, collect historical operating parameter data of the rotary tiller and build the final random forest model;

[0011] S4, identifying and classifying the standard real-time operating parameter data of the rotary tiller through the final random forest model to generate abnormal operating parameter data of the rotary tiller;

[0012] S5, determining whether the rotary tiller is in an abnormal operating state according to the abnormal operating parameter data of the rotary tiller;

[0013] S6. When the rotary tiller is in an abnormal operating state, the real-time operating parameter data of the rotary tiller is optimized by an intelligent optimization algorithm to obtain the optimized operating parameter data of the rotary tiller;

[0014] S7. Construct rotary tiller operating parameter control data based on the optimized rotary tiller operating parameter data, and execute the rotary tiller operating parameter intelligent control operation.

[0015] The present invention generates standard real-time operating parameter data of a rotary tiller by performing mean processing on the collected real-time operating parameter data of the rotary tiller, thereby ensuring the accuracy of the acquired data and avoiding the error caused by a single data; constructs a final random forest model by collecting historical operating parameter data of the rotary tiller as a data sample set, provides reliable data for the construction of the random forest model, and ensures the accuracy of the final random forest model; identifies and classifies the standard real-time operating parameter data of the rotary tiller by the final random forest model, generates abnormal operating parameter data of the rotary tiller, realizes rapid identification of abnormal data, and improves the speed and accuracy of identification; calculates the abnormal value of the real-time state of the rotary tiller by counting the number of abnormal operating parameter data of the rotary tiller, and analyzes the numerical value of the abnormal value of the real-time state of the rotary tiller to determine whether the rotary tiller is in an abnormal operating state, thereby ensuring the accuracy of the determination result; optimizes the operating parameters of the rotary tiller by using a multiverse algorithm, has a good optimization effect and a high efficiency of the optimization process; scientifically constructs the control data of the operating parameters of the rotary tiller, efficiently performs the intelligent control operation of the operating parameters of the rotary tiller, and realizes the precise control and adjustment of the real-time operating state of the rotary tiller.

[0016] Preferably, the parameter data collection time point set and the rotary tiller operation parameter category set are set, and the operation parameter data of the rotary tiller during operation is collected in real time according to the parameter data collection time point set and the rotary tiller operation parameter category set, and the specific steps of obtaining the real-time operation parameter data matrix of the rotary tiller are as follows:

[0017] S11, setting a parameter data collection cycle, dividing the parameter data collection cycle evenly to obtain a parameter data collection time point set ,in, Indicates Parameter data collection time point, Indicates the total number of parameter data collection time points;

[0018] S12. Set the rotary tiller operating parameter category set ,in, Indicates Rotary tiller operating parameters: Indicates the total number of rotary tiller operating parameter categories;

[0019] S13, collecting time point set according to the parameter data And the rotary tiller operating parameter category set The operating parameter data of the rotary tiller during operation is collected in real time to obtain the real-time operating parameter data matrix of the rotary tiller. as follows:

[0020]

[0021] in, Indicates that the rotary tiller is The first parameter data collection time point Class real-time operation parameter data.

[0022] Preferably, the specific steps of numerically processing the real-time operating parameter data of the rotary tiller in the real-time operating parameter data matrix to generate standard real-time operating parameter data of the rotary tiller are as follows:

[0023] S21, using the mean formula to calculate the real-time operating parameter data matrix of the rotary tiller The real-time operating parameter data of the rotary tiller in the system are processed by average numerical measurement to generate a standard real-time operating parameter data set of the rotary tiller. ,in, Indicates Standard real-time operating parameter data of rotary tiller operating parameters, and .

[0024] The present invention generates standard real-time operating parameter data of the rotary tiller by performing mean processing on the collected real-time operating parameter data of the rotary tiller, thereby ensuring the accuracy of the acquired data and avoiding errors caused by single data.

[0025] Preferably, the specific steps of collecting the historical operating parameter data of the rotary tiller and constructing the final random forest model are as follows:

[0026] S31, collecting historical operating parameter data of the rotary tiller as a data sample set;

[0027] S32, set the training data ratio to And the test data ratio is , dividing the data sample set according to the training data ratio and the test data ratio to obtain a training data sample set and a test data sample set respectively;

[0028] S33, adopting the unweighted sampling method to perform sampling on the training data sample set Random sampling, we get training data sample subsets, and the number of training data samples in each training data sample subset is the same;

[0029] S34. Construction Initial decision tree models are trained for each training data sample subset, and an initial decision tree model is obtained. Trained decision tree models are combined to form the initial random forest model. The training process of the decision tree is as follows:

[0030] S341, taking each training data sample in the training data sample subset as the root node of the initial decision tree model;

[0031] S342, using the C4.5 algorithm to split the root node of the initial decision tree model to generate child nodes, and then using the generated child nodes as the root nodes to split again until no new child nodes are generated, the operation is terminated, and a trained decision tree model is obtained; the root node splitting process is as follows:

[0032] S3421, setting the information gain rate threshold;

[0033] S3422, calculating the information gain rate of each rotary tiller historical operating parameter data in the root node of the initial decision tree model;

[0034] If any calculated information gain rate is greater than or equal to the information gain rate threshold, all calculated information gain rates are arranged from large to small, and the root node of the initial decision tree model is split according to the historical operating parameter data of the rotary tiller with the largest information gain rate;

[0035] If all calculated information gain rates are less than the information gain rate threshold, the node splitting is stopped;

[0036] S35. Set the accuracy threshold of the initial random forest model, test the initial random forest model through the test data sample set and calculate the accuracy of the test result. If the accuracy of the test result is greater than or equal to the accuracy threshold of the initial random forest model, the initial random forest model is used as the final random forest model; otherwise, adjust the parameters of the initial random forest model and repeat the operations in S33 to S35 until the calculated accuracy of the test result is greater than or equal to the accuracy threshold of the initial random forest model.

[0037] By collecting the historical operating parameter data of the rotary tiller as the data sample set, the final random forest model was constructed, which provided reliable data for the construction of the random forest model and ensured the accuracy of the final random forest model;

[0038] Preferably, the specific steps of identifying and classifying the standard real-time operating parameter data of the rotary tiller by the final random forest model and generating abnormal operating parameter data of the rotary tiller are as follows:

[0039] S41, inputting the standard real-time operating parameter data of the rotary tiller in the standard real-time operating parameter data set into the final random forest model for identification and classification, filtering out abnormal data in the standard real-time operating parameter data of the rotary tiller according to the classification result, and generating abnormal operating parameter data of the rotary tiller.

[0040] The standard real-time operating parameter data of the rotary tiller is identified and classified through the final random forest model to generate abnormal operating parameter data of the rotary tiller, thereby realizing rapid identification of abnormal data and improving the speed and accuracy of identification.

[0041] Preferably, the specific steps of determining whether the rotary tiller is in an abnormal operating state according to the abnormal operating parameter data of the rotary tiller are as follows:

[0042] S51, setting the abnormal state threshold of the rotary tiller to ;

[0043] S52: Counting the number of abnormal operation parameter data of the rotary tiller , calculate the real-time abnormal value of the rotary tiller ;

[0044] S53, by analyzing the abnormal value of the real-time state of the rotary tiller The value of determines whether the rotary tiller is in an abnormal operating state;

[0045] when When the rotary tiller is judged to be in normal operation, no operation is performed;

[0046] when When the rotary tiller is in normal operation, the rotary tiller is returned to S1 to recalculate the abnormal value of the real-time state of the rotary tiller. If the recalculated abnormal value of the real-time state of the rotary tiller is 0, the rotary tiller is determined to be in normal operation; otherwise, the rotary tiller is determined to be in abnormal operation;

[0047] when When the rotary tiller is in abnormal operation state.

[0048] By counting the number of abnormal operating parameter data of the rotary tiller, calculating the real-time state abnormal value of the rotary tiller, and analyzing the numerical value of the real-time state abnormal value of the rotary tiller, it is determined whether the rotary tiller is in an abnormal operating state, thereby ensuring the accuracy of the determination result.

[0049] Preferably, when the rotary tiller is in an abnormal operating state, the real-time operating parameter data of the rotary tiller is optimized by an intelligent optimization algorithm, and the specific steps of obtaining the optimized operating parameter data of the rotary tiller are as follows:

[0050] S61. Construct a multiverse population and set the multiverse population size to , the current number of iterations is The maximum number of iterations is And the search space dimension of rotary tiller operation parameter data is ;

[0051] Randomly generate the standard real-time operating parameter data of the rotary tiller according to the standard real-time operating parameter data set of the rotary tiller A set of rotary tiller operating parameter data, each multiverse individual in the multiverse population corresponds to a set of rotary tiller operating parameter data;

[0052] S62. Calculate the expansion rate of each multiverse individual in the multiverse population, and take the multiverse individual with the highest expansion rate as the current optimal individual; the expansion rate calculation formula is as follows:

[0053]

[0054] in, Indicates the first The expansion rate of each multiverse entity, Indicates the first The working efficiency of the rotary tiller under the rotary tiller operating parameter data corresponding to the multiverse individuals, Indicates the correction value;

[0055] S63. There are white holes and black holes in each multiverse individual. The operating parameter data of each rotary tiller in the multiverse individual is exchanged with the corresponding operating parameter data of the rotary tiller in other multiverse individuals through the white holes and black holes to complete its own update. The exchange formula is as follows:

[0056]

[0057] in, Indicates of the multiverse Rotary tiller operating parameter data, represents a random number between [0,1], Indicates The normalized expansion rate of each multiverse entity, Indicates the first of the multiverse Rotavator-like operating parameter data;

[0058] S64. Each multiverse individual in the multiverse population always establishes a wormhole tunnel with the current optimal individual. Each rotary tiller operating parameter data in any multiverse individual can be updated to the corresponding rotary tiller operating parameter data in the current optimal individual through the wormhole tunnel. The update formula is as follows:

[0059]

[0060] in, represents the best individual among the current Rotary tiller operating parameter data, and Respectively represent The upper and lower bounds of the rotary tiller operating parameter data in the rotary tiller operating parameter data search space, , and Both represent random numbers between [0,1]; represents the probability of the existence of a wormhole, represents the moving step length; the calculation formula for the wormhole existence probability and the moving step length is as follows:

[0061]

[0062]

[0063] in, Indicates the current level of mining of multiverse individuals. and Respectively and of Power;

[0064] S65, recalculate the expansion value of each multiverse individual in the updated multiverse population, arrange the expansion values ​​of each multiverse individual from large to small, select the multiverse individual with the highest expansion value as the updated optimal individual, if the expansion value of the updated optimal individual is greater than or equal to the expansion value of the current optimal individual, replace the current optimal individual with the updated optimal individual; otherwise, retain the current optimal individual;

[0065] S66, determine the current number of iterations Is it less than the maximum number of iterations? , if the current number of iterations Less than the maximum number of iterations , then the current number of iterations Add 1 and return to S63; otherwise, take the current optimal individual as the global optimal solution, and output the rotary tiller operating parameter data corresponding to the global optimal solution as the optimized rotary tiller operating parameter data.

[0066] Compared with the traditional parameter control method, the present invention adopts the multiverse algorithm to optimize the operating parameters of the rotary tiller, has a better optimization effect and a higher efficiency in the optimization process. The operating parameters of the rotary tiller can be optimized according to the real-time feedback data, and the loss caused by the abnormal state of the rotary tiller can be avoided in time, thereby realizing the intelligent control of the parameters of the rotary tiller.

[0067] Preferably, the specific steps of constructing the rotary tiller operating parameter control data based on the optimized rotary tiller operating parameter data and performing the rotary tiller operating parameter intelligent control operation are as follows:

[0068] S71, marking the optimized rotary tiller operating parameter data to generate rotary tiller operating parameter control data And feedback to the main control unit through the wireless communication network;

[0069] S72, the main control unit adjusts the data according to the received rotary tiller operating parameters Perform intelligent control of rotary tiller operating parameters.

[0070] By scientifically constructing the rotary tiller operating parameter control data, the rotary tiller operating parameter intelligent control operation can be efficiently performed to achieve precise control and adjustment of the rotary tiller's real-time operating status.

[0071] The present invention also discloses a rotary tiller parameter intelligent control system based on machine learning, comprising a rotary tiller operating parameter data acquisition module, a rotary tiller operating parameter data processing module, a final random forest construction module, a rotary tiller abnormal operating parameter data identification module, a rotary tiller abnormal operating state determination module and a rotary tiller operating parameter tuning module;

[0072] The rotary tiller operation parameter data acquisition module collects the operation parameter data of the rotary tiller in real time during the operation of the rotary tiller through the set parameter data acquisition time point set and the set rotary tiller operation parameter category set, and obtains the real-time operation parameter data matrix of the rotary tiller;

[0073] The rotary tiller operation parameter data processing module performs average numerical value measurement processing on the real-time operation parameter data of the rotary tiller through a mean value formula to generate standard real-time operation parameter data of the rotary tiller;

[0074] The final random forest construction module collects historical operating parameter data of the rotary tiller as a data sample set, and performs training and testing based on the data sample set to obtain a final random forest model;

[0075] The rotary tiller abnormal operation parameter data identification module inputs the rotary tiller standard real-time operation parameter data into the final random forest model for identification and classification, and screens out abnormal data in the rotary tiller standard real-time operation parameter data according to the classification result to generate the rotary tiller abnormal operation parameter data;

[0076] The rotary tiller abnormal operation state determination module calculates the real-time state abnormal value of the rotary tiller by counting the number of the abnormal operation parameter data of the rotary tiller, and analyzes the numerical value of the real-time state abnormal value of the rotary tiller to determine whether the rotary tiller is in an abnormal operation state;

[0077] The rotary tiller operating parameter tuning module tunes the real-time operating parameter data of the rotary tiller through a multiverse algorithm to obtain the tuned operating parameter data of the rotary tiller.

[0078] (III) Beneficial effects

[0079] 1. The present invention generates standard real-time operating parameter data of a rotary tiller by performing mean processing on the collected real-time operating parameter data of the rotary tiller, thereby ensuring the accuracy of the acquired data and avoiding the error caused by a single data; by collecting the historical operating parameter data of the rotary tiller as a data sample set, a final random forest model is constructed, which provides reliable data for the construction of the random forest model and ensures the accuracy of the final random forest model; the standard real-time operating parameter data of the rotary tiller is identified and classified by the final random forest model to generate abnormal operating parameter data of the rotary tiller, thereby realizing rapid identification of abnormal data and improving the speed and accuracy of identification; by counting the number of abnormal operating parameter data of the rotary tiller, calculating the abnormal value of the real-time state of the rotary tiller, and analyzing the numerical value of the abnormal value of the real-time state of the rotary tiller to determine whether the rotary tiller is in an abnormal operating state, thereby ensuring the accuracy of the determination result;

[0080] 2. Compared with the traditional parameter control method, the present invention adopts the multiverse algorithm to optimize the operating parameters of the rotary tiller, which has a better optimization effect and a higher efficiency of the optimization process. The operating parameters of the rotary tiller can be optimized according to the real-time feedback data, and the loss caused by the abnormal state of the rotary tiller can be avoided in time, so as to realize the intelligent control of the parameters of the rotary tiller. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0082] Figure 1 A flow chart of a method for intelligently controlling parameters of a rotary tiller based on machine learning provided by the present invention;

[0083] Figure 2 A module schematic diagram of a rotary tiller parameter intelligent control system based on machine learning provided by the present invention. DETAILED DESCRIPTION

[0084] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0085] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0086] Embodiment 1 is as follows:

[0087] See also Figure 1 , a method for intelligently controlling parameters of a rotary tiller based on machine learning, comprising the following steps:

[0088] S1, setting a parameter data collection time point set and a rotary tiller operation parameter category set, and collecting the operation parameter data of the rotary tiller in real time during the operation process according to the parameter data collection time point set and the rotary tiller operation parameter category set to obtain a real-time operation parameter data matrix of the rotary tiller;

[0089] S11, setting a parameter data collection cycle, dividing the parameter data collection cycle evenly to obtain a parameter data collection time point set ,in, Indicates Parameter data collection time point, Indicates the total number of parameter data collection time points;

[0090] S12. Set the rotary tiller operating parameter category set ,in, Indicates Rotary tiller operating parameters: Indicates the total number of rotary tiller operating parameter categories;

[0091] S13, collecting time point set according to the parameter data And the rotary tiller operating parameter category set The operating parameter data of the rotary tiller during operation is collected in real time to obtain the real-time operating parameter data matrix of the rotary tiller. as follows:

[0092]

[0093] in, Indicates that the rotary tiller is The first parameter data collection time point Class real-time operation parameter data.

[0094] S2, performing numerical processing on the real-time operating parameter data of the rotary tiller in the real-time operating parameter data matrix of the rotary tiller to generate standard real-time operating parameter data of the rotary tiller;

[0095] S21, using the mean formula to calculate the real-time operating parameter data matrix of the rotary tiller The real-time operating parameter data of the rotary tiller in the system are processed by average numerical measurement to generate a standard real-time operating parameter data set of the rotary tiller. ,in, Indicates Standard real-time operating parameter data of rotary tiller operating parameters, and .

[0096] S3, collect historical operating parameter data of the rotary tiller and build the final random forest model;

[0097] S31, collecting historical operating parameter data of the rotary tiller as a data sample set;

[0098] S32, set the training data ratio to And the test data ratio is , dividing the data sample set according to the training data ratio and the test data ratio to obtain a training data sample set and a test data sample set respectively;

[0099] S33, adopting the unweighted sampling method to perform sampling on the training data sample set Random sampling, we get training data sample subsets, and the number of training data samples in each training data sample subset is the same;

[0100] S34. Construction Initial decision tree models are trained for each training data sample subset, and an initial decision tree model is obtained. Trained decision tree models are combined to form the initial random forest model. The training process of the decision tree is as follows:

[0101] S341, taking each training data sample in the training data sample subset as the root node of the initial decision tree model;

[0102] S342, using the C4.5 algorithm to split the root node of the initial decision tree model to generate child nodes, and then using the generated child nodes as the root nodes to split again until no new child nodes are generated, the operation is terminated, and a trained decision tree model is obtained; the root node splitting process is as follows:

[0103] S3421, setting the information gain rate threshold;

[0104] S3422, calculating the information gain rate of each rotary tiller historical operating parameter data in the root node of the initial decision tree model;

[0105] If any calculated information gain rate is greater than or equal to the information gain rate threshold, all calculated information gain rates are arranged from large to small, and the root node of the initial decision tree model is split according to the historical operating parameter data of the rotary tiller with the largest information gain rate;

[0106] If all calculated information gain rates are less than the information gain rate threshold, the node splitting is stopped;

[0107] S35. Set the accuracy threshold of the initial random forest model, test the initial random forest model through the test data sample set and calculate the accuracy of the test result. If the accuracy of the test result is greater than or equal to the accuracy threshold of the initial random forest model, the initial random forest model is used as the final random forest model; otherwise, adjust the parameters of the initial random forest model and repeat the operations in S33 to S35 until the calculated accuracy of the test result is greater than or equal to the accuracy threshold of the initial random forest model.

[0108] S4, identifying and classifying the standard real-time operating parameter data of the rotary tiller through the final random forest model to generate abnormal operating parameter data of the rotary tiller;

[0109] S41, inputting the standard real-time operating parameter data of the rotary tiller in the standard real-time operating parameter data set into the final random forest model for identification and classification, filtering out abnormal data in the standard real-time operating parameter data of the rotary tiller according to the classification result, and generating abnormal operating parameter data of the rotary tiller.

[0110] S5, determining whether the rotary tiller is in an abnormal operating state according to the abnormal operating parameter data of the rotary tiller;

[0111] S51, setting the abnormal state threshold of the rotary tiller to ;

[0112] S52: Counting the number of abnormal operation parameter data of the rotary tiller , calculate the real-time abnormal value of the rotary tiller ;

[0113] S53, by analyzing the abnormal value of the real-time state of the rotary tiller The value of determines whether the rotary tiller is in an abnormal operating state;

[0114] when When the rotary tiller is judged to be in normal operation, no operation is performed;

[0115] when When the rotary tiller is in normal operation, the rotary tiller is returned to S1 to recalculate the abnormal value of the real-time state of the rotary tiller. If the recalculated abnormal value of the real-time state of the rotary tiller is 0, the rotary tiller is determined to be in normal operation; otherwise, the rotary tiller is determined to be in abnormal operation;

[0116] when When the rotary tiller is in abnormal operation state.

[0117] S6. When the rotary tiller is in an abnormal operating state, the real-time operating parameter data of the rotary tiller is optimized by an intelligent optimization algorithm to obtain the optimized operating parameter data of the rotary tiller;

[0118] S61. Construct a multiverse population and set the multiverse population size to , the current number of iterations is The maximum number of iterations is And the search space dimension of rotary tiller operation parameter data is ;

[0119] Randomly generate the standard real-time operating parameter data of the rotary tiller according to the standard real-time operating parameter data set of the rotary tiller A set of rotary tiller operating parameter data, each multiverse individual in the multiverse population corresponds to a set of rotary tiller operating parameter data;

[0120] S62. Calculate the expansion rate of each multiverse individual in the multiverse population, and take the multiverse individual with the highest expansion rate as the current optimal individual; the expansion rate calculation formula is as follows:

[0121]

[0122] in, Indicates the first The expansion rate of each multiverse entity, Indicates the first The working efficiency of the rotary tiller under the rotary tiller operating parameter data corresponding to the multiverse individuals, Indicates the correction value;

[0123] S63. There are white holes and black holes in each multiverse individual. The operating parameter data of each rotary tiller in the multiverse individual is exchanged with the corresponding operating parameter data of the rotary tiller in other multiverse individuals through the white holes and black holes to complete its own update. The exchange formula is as follows:

[0124]

[0125] in, Indicates of the multiverse Rotary tiller operating parameter data, represents a random number between [0,1], Indicates The normalized expansion rate of each multiverse entity, Indicates the first of the multiverse Rotavator-like operating parameter data;

[0126] S64. Each multiverse individual in the multiverse population always establishes a wormhole tunnel with the current optimal individual. Each rotary tiller operating parameter data in any multiverse individual can be updated to the corresponding rotary tiller operating parameter data in the current optimal individual through the wormhole tunnel. The update formula is as follows:

[0127]

[0128] in, represents the best individual among the current Rotary tiller operating parameter data, and Respectively represent The upper and lower bounds of the rotary tiller operating parameter data in the rotary tiller operating parameter data search space, , and Both represent random numbers between [0,1]; represents the probability of the existence of a wormhole, represents the moving step length; the calculation formula for the wormhole existence probability and the moving step length is as follows:

[0129]

[0130]

[0131] in, Indicates the current level of mining of multiverse individuals. and Respectively and of Power;

[0132] S65, recalculate the expansion value of each multiverse individual in the updated multiverse population, arrange the expansion values ​​of each multiverse individual from large to small, select the multiverse individual with the highest expansion value as the updated optimal individual, if the expansion value of the updated optimal individual is greater than or equal to the expansion value of the current optimal individual, replace the current optimal individual with the updated optimal individual; otherwise, retain the current optimal individual;

[0133] S66, determine the current number of iterations Is it less than the maximum number of iterations? , if the current number of iterations Less than the maximum number of iterations , then the current number of iterations Add 1 and return to S63; otherwise, take the current optimal individual as the global optimal solution, and output the rotary tiller operating parameter data corresponding to the global optimal solution as the optimized rotary tiller operating parameter data.

[0134] S7, constructing rotary tiller operating parameter control data based on the optimized rotary tiller operating parameter data, and performing a rotary tiller operating parameter intelligent control operation;

[0135] S71, marking the optimized rotary tiller operating parameter data to generate rotary tiller operating parameter control data And feedback to the main control unit through the wireless communication network;

[0136] S72, the main control unit adjusts the data according to the received rotary tiller operating parameters Perform intelligent control of rotary tiller operating parameters.

[0137] Embodiment 2 is as follows:

[0138] See also Figure 2 , a rotary tiller parameter intelligent control system based on machine learning, including a rotary tiller operating parameter data acquisition module, a rotary tiller operating parameter data processing module, a final random forest construction module, a rotary tiller abnormal operating parameter data recognition module, a rotary tiller abnormal operating state determination module and a rotary tiller operating parameter tuning module;

[0139] The rotary tiller operation parameter data acquisition module collects the operation parameter data of the rotary tiller in real time during the operation of the rotary tiller through the set parameter data acquisition time point set and the set rotary tiller operation parameter category set, and obtains the real-time operation parameter data matrix of the rotary tiller;

[0140] The rotary tiller operation parameter data processing module performs average numerical value measurement processing on the real-time operation parameter data of the rotary tiller through a mean value formula to generate standard real-time operation parameter data of the rotary tiller;

[0141] The final random forest construction module collects historical operating parameter data of the rotary tiller as a data sample set, and performs training and testing based on the data sample set to obtain a final random forest model;

[0142] The rotary tiller abnormal operation parameter data identification module inputs the rotary tiller standard real-time operation parameter data into the final random forest model for identification and classification, and screens out abnormal data in the rotary tiller standard real-time operation parameter data according to the classification result to generate the rotary tiller abnormal operation parameter data;

[0143] The rotary tiller abnormal operation state determination module calculates the real-time state abnormal value of the rotary tiller by counting the number of the abnormal operation parameter data of the rotary tiller, and analyzes the numerical value of the real-time state abnormal value of the rotary tiller to determine whether the rotary tiller is in an abnormal operation state;

[0144] The rotary tiller operating parameter tuning module tunes the real-time operating parameter data of the rotary tiller through a multiverse algorithm to obtain the tuned operating parameter data of the rotary tiller.

[0145] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0146] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A method for intelligently controlling parameters of a rotary tiller based on machine learning, characterized in that: The steps include: S1, setting a parameter data collection time point set and a rotary tiller operation parameter category set, and collecting the operation parameter data of the rotary tiller in real time during the operation process according to the parameter data collection time point set and the rotary tiller operation parameter category set to obtain a real-time operation parameter data matrix of the rotary tiller; S2, performing numerical processing on the real-time operating parameter data of the rotary tiller in the real-time operating parameter data matrix of the rotary tiller to generate standard real-time operating parameter data of the rotary tiller; S3, collect historical operating parameter data of the rotary tiller and build the final random forest model; S4, identifying and classifying the standard real-time operating parameter data of the rotary tiller through the final random forest model to generate abnormal operating parameter data of the rotary tiller; S5, determining whether the rotary tiller is in an abnormal operating state according to the abnormal operating parameter data of the rotary tiller; S6. When the rotary tiller is in an abnormal operating state, the real-time operating parameter data of the rotary tiller is optimized by an intelligent optimization algorithm to obtain the optimized operating parameter data of the rotary tiller; S7. Construct rotary tiller operating parameter control data based on the optimized rotary tiller operating parameter data, and perform rotary tiller operating parameter intelligent control operations.

2. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 1, characterized in that: The S1 comprises the following steps: S11, setting a parameter data collection cycle, dividing the parameter data collection cycle evenly to obtain a parameter data collection time point set ,in, Indicates Parameter data collection time point, Indicates the total number of parameter data collection time points; S12. Set the rotary tiller operating parameter category set ,in, Indicates Rotary tiller operating parameters: Indicates the total number of rotary tiller operating parameter categories; S13, collecting time point set according to the parameter data And the rotary tiller operating parameter category set The operating parameter data of the rotary tiller during operation is collected in real time to obtain the real-time operating parameter data matrix of the rotary tiller. .

3. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 2 is characterized in that: The S2 comprises the following steps: S21, using the mean formula to calculate the real-time operating parameter data matrix of the rotary tiller The real-time operating parameter data of the rotary tiller in the system are processed by average numerical measurement to generate a standard real-time operating parameter data set of the rotary tiller. ,in, Indicates Standard real-time operating parameter data of rotary tiller operating parameters.

4. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 3 is characterized in that: The S3 comprises the following steps: S31, collecting historical operating parameter data of the rotary tiller as a data sample set; S32, set the training data ratio to And the test data ratio is , dividing the data sample set according to the training data ratio and the test data ratio to obtain a training data sample set and a test data sample set respectively; S33, adopting the unweighted sampling method to perform sampling on the training data sample set Random sampling, we get training data sample subsets, and the number of training data samples in each training data sample subset is the same; S34. Construction Initial decision tree models are trained for each training data sample subset, and an initial decision tree model is obtained. Trained decision tree models are combined to form the initial random forest model. The training process of the decision tree is as follows: S341, taking each training data sample in the training data sample subset as the root node of the initial decision tree model; S342, using the C4.5 algorithm to split the root node of the initial decision tree model to generate child nodes, and then using the generated child nodes as the root nodes to split again until no new child nodes are generated, the operation is terminated, and a trained decision tree model is obtained; the root node splitting process is as follows: S3421, setting the information gain rate threshold; S3422, calculating the information gain rate of each rotary tiller historical operating parameter data in the root node of the initial decision tree model; If any calculated information gain rate is greater than or equal to the information gain rate threshold, all calculated information gain rates are arranged from large to small, and the root node of the initial decision tree model is split according to the historical operating parameter data of the rotary tiller with the largest information gain rate; If all calculated information gain rates are less than the information gain rate threshold, the node splitting is stopped; S35. Set the accuracy threshold of the initial random forest model, test the initial random forest model through the test data sample set and calculate the accuracy of the test result. If the accuracy of the test result is greater than or equal to the accuracy threshold of the initial random forest model, the initial random forest model is used as the final random forest model; otherwise, adjust the parameters of the initial random forest model and repeat the operations in S33 to S35 until the calculated accuracy of the test result is greater than or equal to the accuracy threshold of the initial random forest model.

5. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 4 is characterized in that: The S4 comprises the following steps: S41, inputting the standard real-time operating parameter data of the rotary tiller in the standard real-time operating parameter data set into the final random forest model for identification and classification, filtering out abnormal data in the standard real-time operating parameter data of the rotary tiller according to the classification result, and generating abnormal operating parameter data of the rotary tiller.

6. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 5, characterized in that: The S5 comprises the following steps: S51, setting the abnormal state threshold of the rotary tiller to ; S52: Counting the number of abnormal operation parameter data of the rotary tiller , calculate the real-time abnormal value of the rotary tiller ; S53, by analyzing the abnormal value of the real-time state of the rotary tiller The value of can be used to determine whether the rotary tiller is in an abnormal operating state.

7. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 6, characterized in that: The method of determining whether the rotary tiller is in an abnormal operating state by analyzing the numerical value of the abnormal value of the real-time state of the rotary tiller comprises the following steps: when When the rotary tiller is judged to be in normal operation, no operation is performed; when When the rotary tiller is in normal operation, the rotary tiller is returned to S1 to recalculate the abnormal value of the real-time state of the rotary tiller. If the recalculated abnormal value of the real-time state of the rotary tiller is 0, the rotary tiller is determined to be in normal operation; otherwise, the rotary tiller is determined to be in abnormal operation; when When the rotary tiller is in abnormal operation state.

8. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 7, characterized in that: The S6 comprises the following steps: S61. Construct a multiverse population and set the multiverse population size to , the current number of iterations is The maximum number of iterations is And the search space dimension of rotary tiller operation parameter data is ; Randomly generate the standard real-time operating parameter data of the rotary tiller according to the standard real-time operating parameter data set of the rotary tiller A set of rotary tiller operating parameter data, each multiverse individual in the multiverse population corresponds to a set of rotary tiller operating parameter data; S62. Calculate the expansion rate of each multiverse individual in the multiverse population, and take the multiverse individual with the highest expansion rate as the current optimal individual; S63. There are white holes and black holes in each multiverse individual. The operating parameter data of each rotary tiller in the multiverse individual is exchanged with the corresponding operating parameter data of the rotary tiller in other multiverse individuals through the white holes and black holes to complete its own update. S64, each multiverse individual in the multiverse population always establishes a wormhole tunnel with the current optimal individual, and each rotary tiller operating parameter data in any multiverse individual can be updated to the corresponding rotary tiller operating parameter data in the current optimal individual through the wormhole tunnel; S65, recalculate the expansion value of each multiverse individual in the updated multiverse population, arrange the expansion values ​​of each multiverse individual from large to small, select the multiverse individual with the highest expansion value as the updated optimal individual, if the expansion value of the updated optimal individual is greater than or equal to the expansion value of the current optimal individual, replace the current optimal individual with the updated optimal individual; otherwise, retain the current optimal individual; S66, determine the current number of iterations Is it less than the maximum number of iterations? , if the current number of iterations Less than the maximum number of iterations , then the current number of iterations Add 1 and return to S63; otherwise, take the current optimal individual as the global optimal solution, and output the rotary tiller operating parameter data corresponding to the global optimal solution as the optimized rotary tiller operating parameter data.

9. The method for intelligently controlling parameters of a rotary tiller based on machine learning according to claim 8, characterized in that: The S7 comprises the following steps: S71, marking the optimized rotary tiller operating parameter data to generate rotary tiller operating parameter control data And feedback to the main control unit through the wireless communication network; S72, the main control unit adjusts the data according to the received rotary tiller operating parameters Perform intelligent control of rotary tiller operating parameters.

10. A system for implementing the intelligent control method of rotary tiller parameters based on machine learning as described in any one of claims 1 to 9, comprising a rotary tiller operating parameter data acquisition module, a rotary tiller operating parameter data processing module, a final random forest construction module, a rotary tiller abnormal operating parameter data identification module, a rotary tiller abnormal operating state determination module and a rotary tiller operating parameter tuning module.

Images