Enterprise platform operation and maintenance method and system based on machine learning

Abstract

The invention discloses an enterprise platform operation and maintenance method and system based on machine learning, and belongs to the technical field of enterprise platform operation and maintenance and data processing. Operation and maintenance data of an enterprise platform are collected firstly, and the collected operation and maintenance data of the enterprise platform are learned through a machine learning technology; therefore, a rule for analyzing the operation and maintenance data of the enterprise platform can be formed, and then the subsequent operation and maintenance data of the enterprise platform is analyzed depending on the rule, so that real-time monitoring, fault prediction, performance optimization and intelligent decision-making of the operation state of the enterprise platform can be realized, the operation and maintenance efficiency of the enterprise platform is effectively improved, and the operation and maintenance cost is reduced. And stable operation of the platform is ensured.

Description

Technical Field

[0001] The present invention belongs to the field of enterprise platform operation and maintenance technology and data processing technology, and specifically relates to an enterprise platform operation and maintenance method and system based on machine learning. Background Art

[0002] Enterprise operation and maintenance, also known as enterprise IT operation and maintenance (IT Operations), refers to a series of management work and operational activities carried out by the information technology department within an enterprise to ensure the normal operation of IT infrastructure and services. The goal of enterprise operation and maintenance is to ensure the stability, reliability and security of IT systems, while optimizing resource utilization, improving service quality and response speed. With the rapid development of enterprise informatization, enterprise platform operation and maintenance faces many challenges, such as increased system complexity, frequent failures, and insufficient operation and maintenance personnel. Traditional operation and maintenance methods mainly rely on manual experience and are difficult to cope with large-scale, high-concurrency platform environments. Therefore, it is of great significance to study an efficient and intelligent operation and maintenance method. Summary of the invention

[0003] The present invention provides an enterprise platform operation and maintenance method and system based on machine learning to solve the problems existing in the prior art.

[0004] In one aspect, the present invention provides an enterprise platform operation and maintenance method based on machine learning, comprising:

[0005] Receiving historical enterprise platform operation and maintenance data and historical fault types corresponding to the historical enterprise platform operation and maintenance data input by human-computer interaction, obtaining historical sample data, and preprocessing the historical sample data to obtain preprocessed historical sample data;

[0006] Selecting at least one machine learning model from a machine learning model library, and using the preprocessed historical sample data to train the selected at least one machine learning model to obtain at least one trained machine learning model;

[0007] Determine a final machine learning model based on at least one trained machine learning model, and deploy the final machine learning model;

[0008] Collecting real-time enterprise platform operation and maintenance data from the enterprise platform, and using the final machine learning model that has been deployed to identify the real-time enterprise platform operation and maintenance data to determine the real-time fault type; wherein the real-time fault type includes no fault or a specific fault type;

[0009] When the real-time fault type is a specific fault type, a warning message of abnormal operation and maintenance of the enterprise platform is generated, and the warning message is transmitted to the equipment designated by the staff to complete the operation and maintenance of the enterprise platform based on machine learning.

[0010] Furthermore, an enterprise platform operation and maintenance parameter adjustment strategy corresponding to each specific fault type is pre-set;

[0011] When the real-time fault type is a specific fault type, the target enterprise platform operation and maintenance parameter adjustment strategy corresponding to the specific fault type is determined by searching and matching, and the target enterprise platform operation and maintenance parameter adjustment strategy is executed to achieve automatic operation adjustment of the enterprise platform.

[0012] Furthermore, after preprocessing the historical sample data, the preprocessed historical sample data is obtained, including:

[0013] Remove the historical sample data with missing values ​​and outliers to obtain the historical sample data after initial processing;

[0014] Performing character conversion on non-numeric data in the historical sample data after the initial processing to convert all data into numeric data, thereby obtaining the historical sample data after the character conversion;

[0015] The historical sample data after character conversion is normalized to obtain the normalized historical sample data, and the normalized historical sample data is used as the pre-processed historical sample data.

[0016] Furthermore, the machine learning model library is a pre-set database and includes at least an LSTM machine learning model, a CNN machine learning model, a BP machine learning model and an SVM machine learning model.

[0017] Further, the preprocessed historical sample data is used to train at least one selected machine learning model to obtain at least one trained machine learning model, including:

[0018] For any machine learning model, initialize the model parameters of the machine learning model, and group all the model parameters into a particle, repeatedly obtain multiple particles, and complete the initialization of the particle swarm;

[0019] The fitness of each particle is obtained by using the preprocessed historical sample data, and the optimal particle in the particle swarm is obtained according to the fitness of each particle.

[0020] An information interaction algorithm with oscillation terms is used to quickly update each particle in the particle swarm to obtain the quickly updated particles.

[0021] When the fitness corresponding to the particle after the rapid update is greater than the fitness corresponding to the optimal particle, the particle with greater fitness is replaced by the optimal particle;

[0022] Determine whether the total number of times the optimal particle has not been replaced reaches the preset training number threshold K. If so, perform a global optimal search on the optimal particle to obtain the updated optimal particle. Otherwise, perform a global jump search on the optimal particle to obtain the updated optimal particle.

[0023] Determine whether the current number of training times has reached the maximum number of training times. If so, redetermine the optimal particle based on the particles after the rapid update and the optimal particle after the update, and use the parameters in the re-determined optimal particle as the final parameters of the machine learning model to obtain the machine learning model after training. Otherwise, return to the rapid update step;

[0024] Traverse all selected machine learning models and obtain at least one trained machine learning model.

[0025] Furthermore, the fitness of each particle is obtained by using the preprocessed historical sample data, and the optimal particle in the particle swarm is obtained according to the fitness of each particle, including:

[0026] For any particle, the model parameters contained in the particle are applied to the machine learning model, and the historical enterprise platform operation and maintenance data in the preprocessed historical sample data is used as the input of the machine learning model, and the historical fault type in the preprocessed historical sample data is used as the expected label value to obtain the error function value of the machine learning model;

[0027] The error function value of the machine learning model is added to a preset constant term and then the inverse is taken to obtain the fitness corresponding to the particle;

[0028] According to the fitness corresponding to each particle, the particle with the largest fitness is determined as the optimal particle.

[0029] Furthermore, an information interaction algorithm with an oscillation term is used to quickly update each particle in the particle swarm, and the particles after the rapid update are obtained, including:

[0030] For any particle, randomly match it with a particle with the shortest Euclidean distance and greater fitness, and obtain the target following particle corresponding to the particle;

[0031] Obtain the Euclidean distance between a particle and its corresponding target follower particle, and determine the oscillation term that changes adaptively with the number of training times according to the current number of training times and the Euclidean distance;

[0032] Determine the difference term between the particle and the optimal particle, and obtain the first update term corresponding to the particle after limiting it with the first random coefficient and the interaction coefficient;

[0033] Determine the difference term between the particle and the target follower particle, and after limiting it with the second random coefficient and the gravity coefficient, obtain the second update term corresponding to the particle;

[0034] The particles are quickly updated according to the first update item and the second update item to obtain particles after the quick update.

[0035] Furthermore, a global optimal search is performed on the optimal particle to obtain the updated optimal particle, including:

[0036] The third update item corresponding to the optimal particle is determined by using a random distribution vector and an exponential function affected by the current training times;

[0037] The historical memory strategy is used to memorize the optimal particle and determine the fourth update item corresponding to the optimal particle;

[0038] According to the third update item and the fourth update item, a global optimal search is performed on the optimal particle to obtain the optimal particle after fast updating.

[0039] Furthermore, a global jump search is performed on the optimal particle to obtain the updated optimal particle, including:

[0040] Among the particles other than the optimal particle, a first random particle is randomly determined from 20% of the particles with the largest fitness, and a second random particle is randomly determined from the other 80% of the particles;

[0041] A global jump search is performed on the optimal particle according to the first random particle and the second random particle to obtain an updated optimal particle.

[0042] On the other hand, the present invention provides an enterprise platform operation and maintenance system based on machine learning, including: a data acquisition module, a model selection and training module, a model deployment module, a model application module and a warning module;

[0043] The data acquisition module is used to receive historical enterprise platform operation and maintenance data and historical fault types corresponding to the historical enterprise platform operation and maintenance data input by human-computer interaction, obtain historical sample data, and pre-process the historical sample data to obtain pre-processed historical sample data;

[0044] The model selection and training module is used to select at least one machine learning model from the machine learning model library, and use the pre-processed historical sample data to train the selected at least one machine learning model to obtain at least one trained machine learning model;

[0045] The model deployment module is used to determine a final machine learning model according to at least one trained machine learning model, and deploy the final machine learning model;

[0046] The model application module is used to collect real-time enterprise platform operation and maintenance data from the enterprise platform, and use the final machine learning model that has been deployed to identify the real-time enterprise platform operation and maintenance data to determine the real-time fault type; wherein the real-time fault type includes no fault or a specific fault type;

[0047] The warning module is used to generate a warning message of abnormal enterprise platform operation and maintenance when the real-time fault type is a specific fault type, and transmit the warning message to a device designated by the staff to complete the enterprise platform operation and maintenance based on machine learning.

[0048] The present invention provides a machine learning-based enterprise platform operation and maintenance method and system, which first collects the operation and maintenance data of the enterprise platform, and then learns the collected enterprise platform operation and maintenance data through machine learning technology, so as to form a rule for analyzing the enterprise platform operation and maintenance data, and then relies on the rule to analyze the subsequent enterprise platform operation and maintenance data, so as to realize real-time monitoring of the enterprise platform operation status, fault prediction, performance optimization and intelligent decision-making, effectively improve the enterprise platform operation and maintenance efficiency, reduce the operation and maintenance costs, and ensure the stable operation of the platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0050] Figure 1 A flowchart of an enterprise platform operation and maintenance method based on machine learning provided in an embodiment of the present invention.

[0051] Figure 2 A schematic diagram of the structure of an enterprise platform operation and maintenance system based on machine learning provided in an embodiment of the present invention.

[0052] Among them, 201-data acquisition module, 202-model selection and training module, 203-model deployment module, 204-model application module, and 205-warning module.

[0053] The above drawings have shown clear embodiments of the present invention, which will be described in more detail below. These drawings and text descriptions are not intended to limit the scope of the present invention in any way, but to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0054] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0055] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0056] like Figure 1 As shown, an embodiment of the present invention provides an enterprise platform operation and maintenance method based on machine learning, including:

[0057] S101, receiving historical enterprise platform operation and maintenance data and historical fault types corresponding to the historical enterprise platform operation and maintenance data input by human-computer interaction, obtaining historical sample data, and preprocessing the historical sample data to obtain preprocessed historical sample data;

[0058] Operation and maintenance data can be data at a specified location in the system log, performance indicators of the enterprise platform, and user operation records, etc. It is also worth noting that the data length of each data collection needs to be the same to ensure that the machine learning model can correctly identify the data. In the case where a type of data may have different lengths, zero padding can be performed before the short data to ensure that the data length is the same and the data at the same location belongs to the same type of data. Other methods can also be used to ensure the accuracy of data length and data location.

[0059] S102, selecting at least one machine learning model from a machine learning model library, and using the preprocessed historical sample data to train the selected at least one machine learning model to obtain at least one trained machine learning model;

[0060] One or more machine learning models can be set in the machine learning model library, and users can also select one or more machine learning models during the training process to determine the optimal machine learning model.

[0061] S103. Determine a final machine learning model according to at least one trained machine learning model, and deploy the final machine learning model;

[0062] Determining a final machine learning model based on at least one trained machine learning model may include: taking out some historical sample data, and using the taken out part of the historical sample data to test the performance indicators of at least one trained machine learning model, and determining the machine learning model with the best performance indicators as the final machine learning model, thereby ensuring the performance of the machine learning model.

[0063] S104, collecting real-time enterprise platform operation and maintenance data from the enterprise platform, and using the final machine learning model that has been deployed to identify the real-time enterprise platform operation and maintenance data to determine the real-time fault type; wherein the real-time fault type includes no fault or a specific fault type;

[0064] The specific fault type may be various common types, which are not limited in the embodiment of the present invention, such as: enterprise platform being attacked, security loopholes, data leakage, backup failure, etc.

[0065] Optionally, the real-time fault type may also include no fault or possible fault. When a fault is likely to exist, a warning message is generated to enable the staff to handle the exception, thereby ensuring the security of the enterprise platform.

[0066] S105. When the real-time fault type is a specific fault type, a warning message indicating abnormal operation and maintenance of the enterprise platform is generated, and the warning message is transmitted to a device designated by the staff to complete the operation and maintenance of the enterprise platform based on machine learning.

[0067] The present invention provides an enterprise platform operation and maintenance method based on machine learning. The method first collects the operation and maintenance data of the enterprise platform, and then learns the collected enterprise platform operation and maintenance data through machine learning technology, so as to form a rule for analyzing the enterprise platform operation and maintenance data. Then, the subsequent enterprise platform operation and maintenance data are analyzed based on the rule, so as to realize real-time monitoring of the operation status of the enterprise platform, fault prediction, performance optimization and intelligent decision-making, effectively improve the operation and maintenance efficiency of the enterprise platform, reduce the operation and maintenance costs, and ensure the stable operation of the platform.

[0068] In the embodiment of the present invention, an enterprise platform operation and maintenance parameter adjustment strategy corresponding to each specific fault type is preset;

[0069] When the real-time fault type is a specific fault type, the target enterprise platform operation and maintenance parameter adjustment strategy corresponding to the specific fault type is determined by searching and matching, and the target enterprise platform operation and maintenance parameter adjustment strategy is executed to achieve automatic operation adjustment of the enterprise platform.

[0070] In the embodiment of the present invention, after preprocessing the historical sample data, obtaining the preprocessed historical sample data includes:

[0071] Remove the historical sample data with missing values ​​and outliers to obtain the historical sample data after initial processing;

[0072] Performing character conversion on non-numeric data in the historical sample data after the initial processing to convert all data into numeric data, thereby obtaining the historical sample data after the character conversion;

[0073] The historical sample data after character conversion is normalized to obtain the normalized historical sample data, and the normalized historical sample data is used as the pre-processed historical sample data.

[0074] In an embodiment of the present invention, the machine learning model library is a pre-set database and at least includes an LSTM (Long Short-Term Memory) machine learning model, a CNN (Convolutional Neural Network) machine learning model, a BP (Back Propagation) machine learning model and an SVM (Support Vector Machine) machine learning model.

[0075] It is worth noting that it is also possible to set only one machine learning model. For example, only a CNN machine learning model is used for operation and maintenance management, and the weights and bias parameters of the CNN machine learning model are trained, so that a trained machine learning model can be obtained.

[0076] In an embodiment of the present invention, the preprocessed historical sample data is used to train at least one selected machine learning model to obtain at least one trained machine learning model, including:

[0077] For any machine learning model, initialize the model parameters of the machine learning model, and group all the model parameters into a particle, repeatedly obtain multiple particles, and complete the initialization of the particle swarm;

[0078] The fitness of each particle is obtained by using the preprocessed historical sample data, and the optimal particle in the particle swarm is obtained according to the fitness of each particle. For example, the error function value (root mean square error or cross entropy error) corresponding to the particle can be obtained first, and then the error function value is added to a very small constant term (such as 0.001) and the inverse is taken to obtain the fitness of the particle.

[0079] An information interaction algorithm with oscillation terms is used to quickly update each particle in the particle swarm to obtain the quickly updated particles.

[0080] When the fitness corresponding to the particle after the rapid update is greater than the fitness corresponding to the optimal particle, the particle with greater fitness is replaced by the optimal particle;

[0081] Determine whether the total number of times the optimal particle has not been replaced reaches the preset training number threshold K. If so, perform a global optimal search on the optimal particle to obtain the updated optimal particle. Otherwise, perform a global jump search on the optimal particle to obtain the updated optimal particle.

[0082] Determine whether the current number of training times has reached the maximum number of training times. If so, redetermine the optimal particle based on the particles after the rapid update and the optimal particle after the update, and use the parameters in the re-determined optimal particle as the final parameters of the machine learning model to obtain the machine learning model after training. Otherwise, return to the rapid update step;

[0083] Traverse all selected machine learning models and obtain at least one trained machine learning model.

[0084] In the prior art, a particle swarm algorithm is often used to optimize the parameters of a machine learning model. Although the particle swarm algorithm has the characteristic of fast training speed, it is very easy to fall into a local optimum. Therefore, an embodiment of the present invention provides an improved particle swarm algorithm to solve the technical problems existing in the existing particle swarm algorithm.

[0085] In an embodiment of the present invention, the fitness corresponding to each particle is obtained by using the preprocessed historical sample data, and the optimal particle in the particle group is obtained according to the fitness corresponding to each particle, including:

[0086] For any particle, the model parameters contained in the particle are applied to the machine learning model, and the historical enterprise platform operation and maintenance data in the preprocessed historical sample data is used as the input of the machine learning model, and the historical fault type in the preprocessed historical sample data is used as the expected label value to obtain the error function value of the machine learning model;

[0087] The error function value of the machine learning model is added to a preset constant term and then the inverse is taken to obtain the fitness corresponding to the particle;

[0088] According to the fitness corresponding to each particle, the particle with the largest fitness is determined as the optimal particle.

[0089] In an embodiment of the present invention, an information interaction algorithm with an oscillation term is used to quickly update each particle in a particle swarm to obtain a quickly updated particle, including:

[0090] For any particle, randomly match it with a particle with the shortest Euclidean distance and greater fitness, and obtain the target following particle corresponding to the particle;

[0091] Get the Euclidean distance between a particle and its corresponding target follower particle And according to the current number of training times and the Euclidean distance, determine the oscillation term that changes adaptively with the number of training times in, represents the i-th particle in the t-th training process, i=1,2,…,L, L represents the total number of particles in the particle swarm, Represents particles The corresponding target follows the particle, T represents the preset maximum number of training times, represents the first random vector, and The total dimension of the parameters of is the same as that of the particles, Each dimension of is a uniformly distributed random number between [-1,1];

[0092] After determining the difference term between the particle and the optimal particle and limiting it with the first random coefficient r1 and the interaction coefficient β, the first update term corresponding to the particle is obtained as: Among them, r1 represents the first random number between (0,1), β represents the interaction coefficient between (0,2), represents the optimal particle;

[0093] After determining the difference term between the particle and the target follower particle and limiting it with the second random coefficient r2 and the gravitational coefficient α, the second update term corresponding to the particle is obtained as: Where r2 represents the second random number between (0,1), and α represents the gravitational coefficient between (0,2);

[0094] According to the first update item and the second update item, the particle is quickly updated, and the particle after the quick update is obtained as follows: in, Represents the particles after rapid update

[0095] The information interaction algorithm with oscillation terms provided in the embodiment of the present invention can utilize information interaction to enable particles to quickly search with global optimal and local optimal information. At the same time, by setting an adaptive oscillation term, it is not easy to fall into the local optimum in the early stage of the algorithm, and has a higher search accuracy in the later stage of the algorithm, which solves the technical problem of being easy to fall into the local optimum in the prior art, and also ensures the optimization ability and optimization accuracy of the algorithm.

[0096] In the embodiment of the present invention, performing a global optimal search on the optimal particle to obtain the updated optimal particle includes:

[0097] Using a random distribution vector and an exponential function affected by the current number of training times, the third update item corresponding to the optimal particle is determined as: in, represents the second random vector, e represents a natural constant, Represents the third update item corresponding to the optimal particle;

[0098] The historical memory strategy is used to memorize the optimal particle, and the fourth update item corresponding to the optimal particle is determined as: Wherein, r3 represents the third random number between (0,1);

[0099] According to the third update item and the fourth update item, a global optimal search is performed on the optimal particle, and the optimal particle after the fast update is obtained as follows: in, Represents the optimal particle in the t-1th training process.

[0100] The global optimal search for the optimal particle provided by the embodiment of the present invention can enable the optimal particle to search in a known optimal region in the solution space, thereby ensuring the local optimization capability of the algorithm.

[0101] In the embodiment of the present invention, performing a global jump search on the optimal particle to obtain the updated optimal particle includes:

[0102] Among the particles other than the best particle, the first random particle is randomly determined from the 20% particles with the largest fitness. and a second random particle is randomly determined from the other 80% of the particles

[0103] According to the first random particle and the second random particle, a global jump search is performed on the optimal particle, and the updated optimal particle is obtained as follows: Wherein, r4 represents the fourth random number between (0,1);

[0104] The global jump search for the optimal particle provided in the embodiment of the present invention is performed only when the optimal particle has not been replaced for multiple times. At this time, it is proved that the population may fall into a local optimum. Therefore, a crossover operation is performed through random particles, thereby realizing a global search.

[0105] Optionally, random flying particles can be used to further enhance the randomness during the global jump search for the optimal particle, such as in Represents particles generated by Levy flight.

[0106] It is worth noting that in order to further speed up the convergence of the algorithm, a greedy strategy can be used to control each search process, but excessive greed may cause the algorithm to easily fall into the local optimum, so only the global jump search can be controlled. And after each update, the particles can be processed out of bounds to ensure the effectiveness of the algorithm.

[0107] like Figure 2 As shown, the embodiment of the present invention provides an enterprise platform operation and maintenance system based on machine learning, including: a data acquisition module 201, a model selection and training module 202, a model deployment module 203, a model application module 204 and a warning module 205;

[0108] The data acquisition module 201 is used to receive historical enterprise platform operation and maintenance data and historical fault types corresponding to the historical enterprise platform operation and maintenance data input by human-computer interaction, obtain historical sample data, and pre-process the historical sample data to obtain pre-processed historical sample data;

[0109] The model selection and training module 202 is used to select at least one machine learning model from the machine learning model library, and use the pre-processed historical sample data to train the selected at least one machine learning model to obtain at least one trained machine learning model;

[0110] The model deployment module 203 is used to determine a final machine learning model according to at least one trained machine learning model, and deploy the final machine learning model;

[0111] The model application module 204 is used to collect real-time enterprise platform operation and maintenance data from the enterprise platform, and use the final machine learning model that has been deployed to identify the real-time enterprise platform operation and maintenance data to determine the real-time fault type; wherein the real-time fault type includes no fault or a specific fault type;

[0112] The warning module 205 is used to generate a warning message of abnormal enterprise platform operation and maintenance when the real-time fault type is a specific fault type, and transmit the warning message to a device designated by the staff to complete the enterprise platform operation and maintenance based on machine learning.

[0113] An enterprise platform operation and maintenance system based on machine learning provided in an embodiment of the present invention can execute the above-mentioned method and technical solution. Its principles and beneficial effects are similar and will not be repeated here.

[0114] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0115] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0116] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0117] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0118] A person of ordinary skill in the art can understand that all or part of the steps in realizing the above-mentioned facts and methods can be completed by instructing the relevant hardware through a program, and the program involved or the program described can be stored in a computer-readable storage medium. When the program is executed, it includes the following steps: At this time, the corresponding method steps are derived, and the storage medium can be ROM / RAM, a disk, an optical disk, etc.

[0119] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for enterprise platform operation and maintenance based on machine learning, characterized in that: include: Receiving historical enterprise platform operation and maintenance data and historical fault types corresponding to the historical enterprise platform operation and maintenance data input by human-computer interaction, obtaining historical sample data, and preprocessing the historical sample data to obtain preprocessed historical sample data; Selecting at least one machine learning model from a machine learning model library, and using the preprocessed historical sample data to train the selected at least one machine learning model to obtain at least one trained machine learning model; Determine a final machine learning model based on at least one trained machine learning model, and deploy the final machine learning model; Collecting real-time enterprise platform operation and maintenance data from the enterprise platform, and using the final machine learning model that has been deployed to identify the real-time enterprise platform operation and maintenance data to determine the real-time fault type; wherein the real-time fault type includes no fault or a specific fault type; When the real-time fault type is a specific fault type, a warning message of abnormal operation and maintenance of the enterprise platform is generated, and the warning message is transmitted to the equipment designated by the staff to complete the operation and maintenance of the enterprise platform based on machine learning.

2. The enterprise platform operation and maintenance method based on machine learning according to claim 1, characterized in that: Pre-set enterprise platform operation and maintenance parameter adjustment strategies corresponding to each specific fault type; When the real-time fault type is a specific fault type, the target enterprise platform operation and maintenance parameter adjustment strategy corresponding to the specific fault type is determined by searching and matching, and the target enterprise platform operation and maintenance parameter adjustment strategy is executed to achieve automatic operation adjustment of the enterprise platform.

3. The enterprise platform operation and maintenance method based on machine learning according to claim 1, characterized in that: After preprocessing the historical sample data, the preprocessed historical sample data is obtained, including: Remove the historical sample data with missing values ​​and outliers to obtain the historical sample data after initial processing; Performing character conversion on non-numeric data in the historical sample data after the initial processing to convert all data into numeric data, thereby obtaining the historical sample data after the character conversion; The historical sample data after character conversion is normalized to obtain the normalized historical sample data, and the normalized historical sample data is used as the pre-processed historical sample data.

4. The enterprise platform operation and maintenance method based on machine learning according to claim 1, characterized in that: The machine learning model library is a pre-set database and includes at least an LSTM machine learning model, a CNN machine learning model, a BP machine learning model and an SVM machine learning model.

5. The enterprise platform operation and maintenance method based on machine learning according to claim 1, characterized in that: Using the preprocessed historical sample data to train at least one selected machine learning model to obtain at least one trained machine learning model, including: For any machine learning model, initialize the model parameters of the machine learning model, and group all the model parameters into a particle, repeatedly obtain multiple particles, and complete the initialization of the particle swarm; The fitness of each particle is obtained by using the preprocessed historical sample data, and the optimal particle in the particle swarm is obtained according to the fitness of each particle. An information interaction algorithm with oscillation terms is used to quickly update each particle in the particle swarm to obtain the quickly updated particles. When the fitness corresponding to the particle after the rapid update is greater than the fitness corresponding to the optimal particle, the particle with greater fitness is replaced by the optimal particle; Determine whether the total number of times the optimal particle has not been replaced reaches the preset training number threshold K. If so, perform a global optimal search on the optimal particle to obtain the updated optimal particle. Otherwise, perform a global jump search on the optimal particle to obtain the updated optimal particle. Determine whether the current number of training times has reached the maximum number of training times. If so, redetermine the optimal particle based on the particles after the rapid update and the optimal particle after the update, and use the parameters in the re-determined optimal particle as the final parameters of the machine learning model to obtain the machine learning model after training. Otherwise, return to the rapid update step; Traverse all selected machine learning models and obtain at least one trained machine learning model.

6. The enterprise platform operation and maintenance method based on machine learning according to claim 5, characterized in that: The fitness of each particle is obtained by using the preprocessed historical sample data, and the optimal particle in the particle swarm is obtained according to the fitness of each particle, including: For any particle, the model parameters contained in the particle are applied to the machine learning model, and the historical enterprise platform operation and maintenance data in the preprocessed historical sample data is used as the input of the machine learning model, and the historical fault type in the preprocessed historical sample data is used as the expected label value to obtain the error function value of the machine learning model; The error function value of the machine learning model is added to a preset constant term and then the inverse is taken to obtain the fitness corresponding to the particle; According to the fitness corresponding to each particle, the particle with the largest fitness is determined as the optimal particle.

7. The enterprise platform operation and maintenance method based on machine learning according to claim 5, characterized in that: The information interaction algorithm with oscillation term is used to quickly update each particle in the particle swarm, and the particles after rapid update are obtained, including: For any particle, randomly match it with a particle with the shortest Euclidean distance and greater fitness, and obtain the target following particle corresponding to the particle; Obtain the Euclidean distance between a particle and its corresponding target follower particle, and determine the oscillation term that changes adaptively with the number of training times according to the current number of training times and the Euclidean distance; Determine the difference term between the particle and the optimal particle, and obtain the first update term corresponding to the particle after limiting it with the first random coefficient and the interaction coefficient; Determine the difference term between the particle and the target follower particle, and after limiting it with the second random coefficient and the gravity coefficient, obtain the second update term corresponding to the particle; The particles are quickly updated according to the first update item and the second update item to obtain particles after the quick update.

8. The enterprise platform operation and maintenance method based on machine learning according to claim 7, characterized in that: Perform a global optimal search for the optimal particle to obtain the updated optimal particle, including: The third update item corresponding to the optimal particle is determined by using a random distribution vector and an exponential function affected by the current training times; The historical memory strategy is used to memorize the optimal particle and determine the fourth update item corresponding to the optimal particle; According to the third update item and the fourth update item, a global optimal search is performed on the optimal particle to obtain the optimal particle after fast updating.

9. The enterprise platform operation and maintenance method based on machine learning according to claim 8, characterized in that: Perform global jump search on the optimal particle to obtain the updated optimal particle, including: Among the particles other than the optimal particle, a first random particle is randomly determined from 20% of the particles with the largest fitness, and a second random particle is randomly determined from the other 80% of the particles; A global jump search is performed on the optimal particle according to the first random particle and the second random particle to obtain an updated optimal particle.

10. An enterprise platform operation and maintenance system based on machine learning, characterized in that: include: Data acquisition module, model selection and training module, model deployment module, model application module and warning module; The data acquisition module is used to receive historical enterprise platform operation and maintenance data and historical fault types corresponding to the historical enterprise platform operation and maintenance data input by human-computer interaction, obtain historical sample data, and pre-process the historical sample data to obtain pre-processed historical sample data; The model selection and training module is used to select at least one machine learning model from the machine learning model library, and use the pre-processed historical sample data to train the selected at least one machine learning model to obtain at least one trained machine learning model; The model deployment module is used to determine a final machine learning model according to at least one trained machine learning model, and deploy the final machine learning model; The model application module is used to collect real-time enterprise platform operation and maintenance data from the enterprise platform, and use the final machine learning model that has been deployed to identify the real-time enterprise platform operation and maintenance data to determine the real-time fault type; wherein the real-time fault type includes no fault or a specific fault type; The warning module is used to generate a warning message of abnormal enterprise platform operation and maintenance when the real-time fault type is a specific fault type, and transmit the warning message to a device designated by the staff to complete the enterprise platform operation and maintenance based on machine learning.

Images