An operation and maintenance method and system based on machine learning

Abstract

The present invention relates to the field of operation and maintenance technology, and is an operation and maintenance method and system based on machine learning. The specific method comprises: collecting historical order data; establishing a customer relationship-machine learning model to predict the potential market value of quartz device processing orders; performing high-dimensional clustering analysis on the potential market value of quartz device processing orders; calculating and obtaining the customer stability of related customers of different quartz device processing orders, and grading corporate customers according to the customer stability; and executing differentiated operation and maintenance services and personalized marketing plans. The present invention solves the problems in the prior art that corporate customer stability of quartz processing factories is poor, and there is a lack of personalized services for customizing different products for different customers.

Description

Technical Field

[0001] The present invention relates to the field of operation and maintenance technology, and is an operation and maintenance method and system based on machine learning. Background Art

[0002] In data operation and maintenance, machine learning technology has been widely used, especially in customer relationship operation and maintenance. By analyzing and processing customer data with machine learning algorithms, the operation and maintenance team can better understand customer behavior patterns and demand trends. In addition, machine learning can also help the operation and maintenance team identify potential problems and risks, and prevent possible system failures or service interruptions in advance through real-time monitoring and predictive analysis, thereby improving service stability and reliability.

[0003] Enterprise customer relationship operation and maintenance refers to the operation management and maintenance work that enterprises carry out in the process of establishing and maintaining good relationships with customers. This includes understanding and responding to customer needs, regularly communicating and exchanging with customers, providing professional services and support, and continuously monitoring customer satisfaction and demand changes. The corporate customers of quartz processing companies are mainly customers who need customized quartz devices or quartz glass products. Their requirements for products may be more specific and professional. In addition, the customers of quartz processing companies are different from one-time transactions or short-term cooperation in other industries. They often need to establish long-term and stable cooperative relationships. How to improve the intelligence of customer relationship management and the sustainability of technical services is an important problem facing quartz processing companies today.

[0004] In the existing disclosed invention technologies, for example, a Chinese patent with publication number CN117829466A discloses a data-driven operation and maintenance service scheduling method, system, device and medium, the method comprising: constructing a GIS electronic map corresponding to the service area according to the basic information of the operated and maintained enterprise and the operation and maintenance personnel; in response to the operation and maintenance request of the operated and maintained enterprise, creating an operation and maintenance service scheduling work order on the GIS electronic map, and assigning operation and maintenance personnel to the operated and maintained enterprise, the status of the operation and maintenance service scheduling work order including pending dispatch, pending acceptance, accepted, pending acceptance, completed and terminated; in response to the operation and maintenance progress feedback from the operation and maintenance personnel, updating the status of the operation and maintenance service scheduling work order; in response to the status of the operation and maintenance service scheduling work order being completed, performing an operation and maintenance service analysis and evaluation on the operation and maintenance personnel, and generating an evaluation result.

[0005] The above-mentioned patent mainly relies on basic information and operation and maintenance requests for service scheduling, which fails to fully consider the customer's personalized needs and customized services. It is not suitable for the enterprise customer relationship operation and maintenance of quartz processing companies, and there is a problem of poor customer service experience. Summary of the invention

[0006] The technical problem to be solved by the present invention is that in the prior art, the corporate customer stability of quartz processing factories is poor, and there is a lack of personalized services for customizing different products for different customers. An operation and maintenance method and system based on machine learning are proposed.

[0007] In order to achieve the above object, the technical solution of the operation and maintenance method based on machine learning of the present invention comprises the following steps:

[0008] S1: Collect historical order data, and perform data preprocessing on the collected historical order data to obtain a processing data sequence and an order data sequence;

[0009] S2: Establish a customer relationship-machine learning model, input the processing data sequence and order data sequence into the customer relationship-machine learning model, and predict the processing order quantity and order total of quartz devices;

[0010] S3: According to the number of processing orders and the total order amount of quartz devices, a cluster analysis model is established to perform high-dimensional cluster analysis on the number of processing orders and the total order amount of quartz devices;

[0011] S4: According to the cluster analysis results and the stability calculation strategy of related customers, the stability of related customers of different quartz device processing orders is calculated, and the enterprise customers are classified according to the stability of the customers to obtain the classification data of the enterprise customers;

[0012] S5: Generate operation and maintenance decision data based on the classification data of enterprise customers.

[0013] Specifically, in S1, the historical order data includes: factory processing data and customer order information data;

[0014] The factory processing data includes: quartz device processing parameter data and factory processing capacity data

[0015] The quartz device processing parameter data includes: quartz device size data and processing process complexity;

[0016] The factory processing capacity data includes: during the quartz device processing, the average number of normal operating equipment in the factory, the average number of workers on duty in the factory, and the factory processing defect rate data;

[0017] The customer order information data includes: customer attention data and customer spending power data

[0018] The customer attention data includes: the average time interval between the signing dates of adjacent customer orders and the number of customer communications and urgings during the quartz device processing;

[0019] The collecting of historical order data and preprocessing of the collected historical order data to obtain a processed data sequence and an order data sequence include: deduplication of the historical order data and processing of missing values; and obtaining a processed data sequence and an order data sequence related to the signing month of the historical order data.

[0020] Specifically, S2 includes the following specific steps:

[0021] S21: Obtain a factory processing data time series and a customer order information data time series, and perform wavelet decomposition on the factory processing data time series and the customer order information data time series;

[0022] When the signing month of the quartz device processing order is t, obtain the periodic subsequence T of the quartz device processing parameter data cs , the trend subsequence R of quartz device processing parameter data cs , random residual subsequence C of quartz device processing parameter data cs ;

[0023] Get the periodic subsequence T of the factory processing capacity data nl , the trend subsequence R of the factory processing capacity data nl , the random residual subsequence C of the factory processing capacity data nl ;

[0024] Get the periodic subsequence T of customer attention data xw , the trend subsequence R of customer attention data xw , random residual subsequence C of customer attention data xw ;

[0025] S22: Establishing a customer relationship-machine learning model, and importing the subsequences obtained by decomposing the factory processing data time series and the customer order information data time series into a sequence correlation fusion strategy to obtain a correlation sequence;

[0026] The sequence correlation fusion strategy is as follows:

[0027]

[0028] The total number of historical order data collected is K, and the historical order data includes processing orders of K quartz devices, where k is a superscript serial number, indicating the processing order of the kth quartz device, and k≤K;

[0029] They are respectively the prediction time series of quartz device processing parameter data, the prediction time series of factory processing capability data, and the prediction time series of customer attention data;

[0030] μ 1，cs , μ 2，cs , μ 3，cs They are the periodic subsequence fusion correlation of quartz device processing parameter data, the trend subsequence fusion correlation of quartz device processing parameter data, and the random residual subsequence fusion correlation of quartz device processing parameter data;

[0031] μ 1，nl , μ 2，nl , μ 3，nl They are the periodic subsequence fusion correlation of the factory processing capacity data, the trend subsequence fusion correlation of the factory processing capacity data, and the random residual subsequence fusion correlation of the factory processing capacity data;

[0032] μ 1，xw , μ 2，xw , μ 3，xw They are the periodic subsequence fusion correlation of customer attention data, the trend subsequence fusion correlation of customer attention data, and the random residual subsequence fusion correlation of customer attention data;

[0033] S23: performing feature selection and feature scaling on the correlation sequence obtained in S22 and inputting the result into the customer relationship-machine learning model, and performing feature compression on the features obtained by feature selection and predicting the potential market value of the processing order of the quartz device;

[0034] The potential market value of the processing orders for the quartz devices is Φ k for:

[0035]

[0036] Among them, α 1 , α 2 , α 3 They are the value factors of quartz device processing parameter data, factory processing capability data, and customer attention data.

[0037] Specifically, in S3, the high-dimensional cluster analysis of the potential market value of the processing orders for quartz devices includes the following specific steps:

[0038] S31: The predicted potential market value of the order is formed into an N-dimensional data feature set L, denoted as L = {a k，n}, where a k，nis the nth characteristic value of the kth quartz device processing order; the data feature set is fuzzy processed by the membership function to obtain the fuzzy processed feature data set H, which is recorded as H = {b k，n}, where b k，n Corresponding to a k，n The characteristic value after fuzzy processing; n represents the number of types of characteristic values;

[0039] S32: According to S31, extract the feature data set H after fuzzy processing, and map the feature data set H to a high-dimensional plane, divide the predicted potential market value of the order into Q categories, and set M cluster center data points, where the cluster center data point set of each potential market value category is G = {g 1 , g 2 , ..., g m}, where the density of feature data points on the high-dimensional plane is ρ; g m is the data point of the mth cluster center; Q = M;

[0040] S33: Construct membership degree d k，m The classification matrix is ​​used to divide and cluster the feature data set H on the high-dimensional plane, where d k，m is the membership degree of the k-th quartz device processing order in the feature data set H to the m-th cluster center data point, where m∈(0,M].

[0041] Specifically, in S4, the calculation strategy of the stability W of the relevant customers is as follows:

[0042]

[0043] Where Z is the number of processing orders for predicted quartz devices;

[0044] is the value level coefficient of the mth cluster center data point; W is the stability of the relevant customers; z is the order number; z≤Z;

[0045] w 0，z The stability of processing order data for the customer’s zth processing order this year;

[0046] The calculation strategy of the stability of the processing order data is as follows:

[0047]

[0048] in, are the predicted time series of quartz device processing parameter data, factory processing capability data, and customer attention data for the zth processing order;

[0049] They are respectively the actual processing parameter data of the quartz device for the zth processing order, the actual factory processing capacity data, and the actual customer attention data.

[0050] Specifically, in S4, the steps of classifying the enterprise customers according to their stability and obtaining the classification data of the enterprise customers are as follows:

[0051] When the stability of relevant customers W≤θ 0 When the customer relationship is unstable, the customer is classified into the customer level;

[0052] When the stability of relevant customers θ 0 <W≤θ 1 When the relevant customers are classified into the general stable relationship customers level;

[0053] When the stability of relevant customers θ 1 <W≤θ 2 When the relevant customers are classified into the level of customers with particularly stable relationships;

[0054] Among them, θ 0 ,θ 1 ,θ 2 are the thresholds for customer stability levels, and θ 0 <θ 1 <θ 1 .

[0055] In addition, the present invention also discloses an operation and maintenance system based on machine learning, which is used to implement the operation and maintenance method based on machine learning, and includes the following modules: a historical data collection module, a sequence prediction module, a cluster analysis module, a customer classification module, and an operation and maintenance decision generation module;

[0056] The historical data collection module is used to collect historical order data, and perform data preprocessing on the collected historical order data to obtain a processing data sequence and an order data sequence;

[0057] The sequence prediction module is used to establish a customer relationship-machine learning model, input the processing data sequence and the order data sequence into the customer relationship-machine learning model, and predict the number of processing orders and the total order amount of quartz devices;

[0058] The cluster analysis module establishes a cluster analysis model according to the number of processing orders and the total order amount of quartz devices, and performs high-dimensional cluster analysis on the number of processing orders and the total order amount of quartz devices;

[0059] The customer classification module is used to calculate the stability of customers related to different quartz device processing orders according to the cluster analysis results and the stability calculation strategy of related customers, and classify the enterprise customers according to the stability of the customers to obtain the classification data of the enterprise customers;

[0060] The operation and maintenance decision generating module generates operation and maintenance decision data according to the classification data of the enterprise customers.

[0061] The present invention also discloses a storage medium, in which instructions are stored. When a computer reads the instructions, the computer executes the operation and maintenance method based on machine learning.

[0062] The present invention also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned operation and maintenance method based on machine learning when executing the computer program.

[0063] Compared with the prior art, the technical effects of the present invention are as follows:

[0064] 1. The present invention takes into account that the customers of quartz processing enterprises are different from one-time transactions or short-term cooperation in other industries, and need to establish long-term and stable cooperative relationships. Through continuous monitoring and feedback mechanisms, differentiated operation and maintenance services and personalized marketing plans are implemented according to the hierarchical data of enterprise customers to improve service quality and customer experience.

[0065] 2. By collecting and preprocessing quartz device processing order data and performing customer relationship analysis based on a machine learning model, the present invention can more scientifically understand customer needs and market trends and improve the accuracy and efficiency of decision-making.

[0066] 3. The present invention can accurately predict the potential market value of quartz device processing orders through the customer relationship-machine learning model and order value clustering analysis model, which helps enterprises to formulate more effective sales strategies and service solutions.

[0067] 4. The present invention uses a customer stability calculation strategy to classify corporate customers according to the stability of customer relationships, which is conducive to maintaining high-quality customer relationships and improving customer loyalty and customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0069] in:

[0070] Figure 1 A flowchart of an operation and maintenance method based on machine learning of the present invention;

[0071] Figure 2 It is a structural schematic diagram of an operation and maintenance system based on machine learning of the present invention. DETAILED DESCRIPTION

[0072] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0073] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0074] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0075] Embodiment 1

[0076] like Figure 1 As shown, an operation and maintenance method based on machine learning disclosed in an embodiment of the present invention includes the following specific steps:

[0077] S1: Collect historical order data, and perform data preprocessing on the collected historical order data to obtain a processing data sequence and an order data sequence;

[0078] In S1, the historical order data includes: factory processing data and customer order information data;

[0079] The factory processing data includes: quartz device processing parameter data and factory processing capacity data

[0080] The quartz device processing parameter data includes: quartz device size data and processing process complexity;

[0081] Exemplarily, in this embodiment, the processing step complexity is the number of steps in the processing of the quartz device.

[0082] The factory processing capacity data includes: during the quartz device processing, the average number of normal operating equipment in the factory, the average number of workers on duty in the factory, and the factory processing defect rate data;

[0083] The customer order information data includes: customer attention data and customer spending power data

[0084] The customer attention data includes: the average time interval between the signing dates of adjacent customer orders and the number of customer communications and urgings during the quartz device processing;

[0085] The collecting of historical order data and preprocessing of the collected historical order data to obtain a processed data sequence and an order data sequence include: deduplication of the historical order data and processing of missing values; and obtaining a processed data sequence and an order data sequence related to the signing month of the historical order data.

[0086] S2: Establish a customer relationship-machine learning model, input the processing data sequence and order data sequence into the customer relationship-machine learning model, and predict the processing order quantity and order total of quartz devices;

[0087] S2 includes the following specific steps:

[0088] S21: Obtain a factory processing data time series and a customer order information data time series, and perform wavelet decomposition on the factory processing data time series and the customer order information data time series;

[0089] When the signing month of the historical order data is t, obtain the periodic subsequence T of the quartz device processing parameter data cs , the trend subsequence R of quartz device processing parameter data cs , random residual subsequence C of quartz device processing parameter data cs ;

[0090] Get the periodic subsequence T of the factory processing capacity data nl , the trend subsequence R of the factory processing capacity data nl , the random residual subsequence C of the factory processing capacity data nl ;

[0091] Get the periodic subsequence T of customer attention data xw , the trend subsequence R of customer attention data xw , random residual subsequence C of customer attention data xw ;

[0092] S22: Establishing a customer relationship-machine learning model, and importing the subsequences obtained by decomposing the factory processing data time series and the customer order information data time series into a sequence correlation fusion strategy to obtain a correlation sequence;

[0093] The sequence correlation fusion strategy is as follows:

[0094]

[0095] The total number of historical order data collected is K, and the historical order data includes K quartz device processing orders, where k is a superscript serial number, indicating the kth quartz device processing order, and k≤K;

[0096] They are respectively the prediction time series of quartz device processing parameter data, the prediction time series of factory processing capability data, and the prediction time series of customer attention data;

[0097] μ 1，cs , μ 2，cs , μ 3，cs They are the periodic subsequence fusion correlation of quartz device processing parameter data, the trend subsequence fusion correlation of quartz device processing parameter data, and the random residual subsequence fusion correlation of quartz device processing parameter data;

[0098] μ 1，nl , μ 2，nl , μ 3，nl They are the periodic subsequence fusion correlation of the factory processing capacity data, the trend subsequence fusion correlation of the factory processing capacity data, and the random residual subsequence fusion correlation of the factory processing capacity data;

[0099] μ 1，xw , μ 2，xw , μ 3，xw They are the periodic subsequence fusion correlation of customer attention data, the trend subsequence fusion correlation of customer attention data, and the random residual subsequence fusion correlation of customer attention data;

[0100] S23: performing feature selection and feature scaling on the correlation sequence obtained in S22 and inputting the result into the customer relationship-machine learning model, and performing feature compression on the features obtained by feature selection and predicting the potential market value of the processing order of the quartz device;

[0101] The potential market value of the processing orders for the quartz devices is Φ k for:

[0102]

[0103] Among them, α 1 , α2 , α 3 They are the value factors of quartz device processing parameter data, factory processing capability data, and customer attention data.

[0104] S3: According to the number of processing orders and the total order amount of quartz devices, a cluster analysis model is established to perform high-dimensional cluster analysis on the number of processing orders and the total order amount of quartz devices;

[0105] In S3, the high-dimensional cluster analysis of the potential market value of the processing orders for quartz devices includes the following specific steps:

[0106] S31: The predicted potential market value of the order is formed into an N-dimensional data feature set L, denoted as L = {a k，n}, where a k，n is the nth characteristic value of the kth quartz device processing order; the data feature set is fuzzy processed by the membership function to obtain the fuzzy processed feature data set H, which is recorded as H = {b k，n}, where b k，n Corresponding to a k，n The characteristic value after fuzzy processing; n represents the number of types of characteristic values;

[0107] S32: According to S31, extract the feature data set H after fuzzy processing, and map the feature data set H to a high-dimensional plane, divide the predicted potential market value of the order into Q categories, and set M cluster center data points, where the cluster center data point set of each potential market value category is G = {g 1 , g 2 , ..., g m}, where the density of feature data points on the high-dimensional plane is ρ; g m is the data point of the mth cluster center; Q = M;

[0108] S33: Construct membership degree d k，m The classification matrix is ​​used to divide and cluster the feature data set H on the high-dimensional plane, where d k，m is the membership degree of the k-th quartz device processing order in the feature data set H to the m-th cluster center data point, where m∈(0,M].

[0109] S4: According to the cluster analysis results and the stability calculation strategy of related customers, the stability of related customers of different quartz device processing orders is calculated, and the enterprise customers are classified according to the stability of the customers to obtain the classification data of the enterprise customers;

[0110] In S4, the calculation strategy of the stability W of the relevant customers is as follows:

[0111]

[0112] Where Z is the number of processing orders for predicted quartz devices;

[0113] is the value level coefficient of the mth cluster center data point; W is the stability of the relevant customers; z is the order number; z≤Z;

[0114] w 0，z The stability of processing order data for the customer’s zth processing order this year;

[0115] The calculation strategy of the stability of the processing order data is as follows:

[0116]

[0117] in, are the predicted time series of quartz device processing parameter data, factory processing capability data, and customer attention data for the zth processing order;

[0118] They are respectively the actual processing parameter data of the quartz device for the zth processing order, the actual factory processing capacity data, and the actual customer attention data.

[0119] In S4, the steps of classifying the enterprise customers according to their stability and obtaining the classification data of the enterprise customers are as follows:

[0120] When the stability of relevant customers W≤θ 0 When the customer relationship is unstable, the customer is classified into the customer level;

[0121] When the stability of relevant customers θ 0 <W≤θ 1 When the relevant customers are classified into the general stable relationship customers level;

[0122] When the stability of relevant customers θ 1 <W≤θ 2 When the relevant customers are classified into the level of customers with particularly stable relationships;

[0123] Among them, θ 0 ,θ 1 ,θ 2 are the thresholds for customer stability levels, and θ 0 <θ 1 <θ 2 .

[0124] S5: Generate operation and maintenance decision data based on the classification data of enterprise customers.

[0125] Exemplarily, in this embodiment, an operation and maintenance decision data in S5 is provided, including: differentiated operation and maintenance services and personalized marketing plans, specifically as follows: accurately push personalized product recommendations, promotional activities and other marketing content according to product characteristics and market positioning, and design customized services and marketing plans according to the characteristics of quartz devices.

[0126] Embodiment 2

[0127] like Figure 2 As shown, an operation and maintenance system based on machine learning disclosed in an embodiment of the present invention includes the following modules:

[0128] Historical data collection module, sequence prediction module, cluster analysis module, customer classification module, and operation and maintenance decision generation module;

[0129] The historical data collection module is used to collect historical order data, and perform data preprocessing on the collected historical order data to obtain a processing data sequence and an order data sequence;

[0130] The sequence prediction module is used to establish a customer relationship-machine learning model, input the processing data sequence and the order data sequence into the customer relationship-machine learning model, and predict the number of processing orders and the total order amount of quartz devices;

[0131] The cluster analysis module establishes a cluster analysis model according to the number of processing orders and the total order amount of quartz devices, and performs high-dimensional cluster analysis on the number of processing orders and the total order amount of quartz devices;

[0132] The customer classification module is used to calculate the stability of customers related to different quartz device processing orders according to the cluster analysis results and the stability calculation strategy of related customers, and classify the enterprise customers according to the stability of the customers to obtain the classification data of the enterprise customers;

[0133] The operation and maintenance decision generating module generates operation and maintenance decision data according to the classification data of the enterprise customers.

[0134] Embodiment 3

[0135] This embodiment provides an electronic device, including: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0136] The processor executes the above-mentioned operation and maintenance method based on machine learning by calling the computer program stored in the memory.

[0137] The electronic device may have relatively large differences due to different configurations or performances, and may include one or more processors (Central Processing Units, CPU) and one or more memories, wherein at least one computer program is stored in the memory, and the computer program is loaded and executed by the processor to implement a machine learning-based operation and maintenance method provided by the above method embodiment. The electronic device may also include other components for implementing device functions, for example, the electronic device may also have components such as a wired or wireless network interface and an input and output interface to input and output data. This embodiment will not be described in detail here.

[0138] Embodiment 4

[0139] This embodiment provides a computer-readable storage medium on which a rewritable computer program is stored.

[0140] When the computer program runs on a computer device, the computer device executes the above-mentioned operation and maintenance method based on machine learning.

[0141] For example, a computer readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0142] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0143] It should be understood that determining B based on A does not mean determining B only based on A. B can also be determined based on A and / or other information.

[0144] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When a computer instruction or computer program is loaded or executed on a computer, a process or function according to an embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or / and a wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0145] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0146] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0148] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0149] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0150] 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 present invention. In this specification, the schematic representations of the above terms do 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.

[0151] In summary, compared with the prior art, the technical effects of the present invention are as follows:

[0152] 1. The present invention takes into account that the customers of quartz processing enterprises are different from one-time transactions or short-term cooperation in other industries, and need to establish long-term and stable cooperative relationships. Through continuous monitoring and feedback mechanisms, differentiated operation and maintenance services and personalized marketing plans are implemented according to the hierarchical data of enterprise customers to improve service quality and customer experience.

[0153] 2. By collecting and preprocessing quartz device processing order data and performing customer relationship analysis based on a machine learning model, the present invention can more scientifically understand customer needs and market trends and improve the accuracy and efficiency of decision-making.

[0154] 3. The present invention can accurately predict the potential market value of quartz device processing orders through the customer relationship-machine learning model and order value clustering analysis model, which helps enterprises to formulate more effective sales strategies and service solutions.

[0155] 4. The present invention uses a customer stability calculation strategy to classify corporate customers according to the stability of customer relationships, which is conducive to maintaining high-quality customer relationships and improving customer loyalty and customer satisfaction.

[0156] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An operation and maintenance method based on machine learning, characterized in that: The method comprises the following specific steps: S1: Collect historical order data, and perform data preprocessing on the collected historical order data to obtain a processing data sequence and an order data sequence; S2: Establish a customer relationship-machine learning model, input the processing data sequence and order data sequence into the customer relationship-machine learning model, and predict the processing order quantity and total order amount of quartz devices; S3: According to the number of processing orders and the total order amount of quartz devices, a cluster analysis model is established to perform high-dimensional cluster analysis on the number of processing orders and the total order amount of quartz devices; S4: According to the cluster analysis results and the stability calculation strategy of related customers, the stability of related customers of different quartz device processing orders is calculated, and the enterprise customers are classified according to the stability of the customers to obtain the classification data of the enterprise customers; In S4, the calculation strategy of the stability W of the relevant customers is as follows: Where Z is the number of processing orders for predicted quartz devices; is the value level coefficient of the mth cluster center data point; W is the stability of the relevant customers; z is the order number; z≤Z; w 0,z The stability of processing order data for the customer’s zth processing order this year; The calculation strategy of the stability of the processing order data is as follows: in, are the predicted time series of quartz device processing parameter data, factory processing capability data, and customer attention data for the zth processing order; They are the actual processing parameter data of the quartz device for the zth processing order, the actual factory processing capacity data, and the actual customer attention data; In S4, the steps of classifying the enterprise customers according to their stability and obtaining the classification data of the enterprise customers are as follows: When the stability of the relevant customers W≤θ0, the relevant customers are classified into the unstable relationship customer level; When the stability of the relevant customers θ0<W≤θ1, the relevant customers are classified as general stable relationship customers; When the stability of the relevant customers θ1<W≤θ2, the relevant customers are classified as particularly stable relationship customers; Among them, θ0, θ1, θ2 are the thresholds for classifying customer stability, and θ0<θ1<θ2; S5: Generate operation and maintenance decision data based on the classification data of enterprise customers.

2. The operation and maintenance method based on machine learning according to claim 1, characterized in that: In S1, the historical order data includes: factory processing data and customer order information data; The factory processing data includes: quartz device processing parameter data and factory processing capacity data The quartz device processing parameter data includes: quartz device size data and processing process complexity; The factory processing capacity data includes: during the quartz device processing, the average number of normal operating equipment in the factory, the average number of workers on duty in the factory, and the factory processing defect rate data; The customer order information data includes: customer attention data and customer spending power data The customer attention data includes: the average time interval between the signing dates of adjacent customer orders and the number of customer communications and urgings during the quartz device processing; The collecting of historical order data and preprocessing of the collected historical order data to obtain a processed data sequence and an order data sequence include: deduplication of the historical order data and processing of missing values; and obtaining a processed data sequence and an order data sequence related to the signing month of the historical order data.

3. The operation and maintenance method based on machine learning according to claim 2, characterized in that: S2 includes the following specific steps: S21: Obtain a factory processing data time series and a customer order information data time series, and perform wavelet decomposition on the factory processing data time series and the customer order information data time series; When the signing month of the historical order data is t, obtain the periodic subsequence T of the quartz device processing parameter data cs , the trend subsequence R of quartz device processing parameter data cs , random residual subsequence C of quartz device processing parameter data cs ; Get the periodic subsequence T of the factory processing capacity data nl , the trend subsequence R of the factory processing capacity data nl , the random residual subsequence C of the factory processing capacity data nl ; Get the periodic subsequence T of customer attention data xw , the trend subsequence R of customer attention data xw , random residual subsequence C of customer attention data xw ; S22: Establishing a customer relationship-machine learning model, and importing the subsequences obtained by decomposing the factory processing data time series and the customer order information data time series into a sequence correlation fusion strategy to obtain a correlation sequence; The sequence correlation fusion strategy is as follows: The total number of historical order data collected is K, and the historical order data includes processing orders of K quartz devices, where k is a superscript serial number, indicating the processing order of the kth quartz device, and k≤K; They are respectively the prediction time series of quartz device processing parameter data, the prediction time series of factory processing capability data and the prediction time series of customer attention data; μ 1,cs ,μ 2,cs ,μ 3,cs They are the periodic subsequence fusion correlation of quartz device processing parameter data, the trend subsequence fusion correlation of quartz device processing parameter data, and the random residual subsequence fusion correlation of quartz device processing parameter data; μ 1,nl ,μ 2,nl ,μ 3,nl They are the periodic subsequence fusion correlation of the factory processing capacity data, the trend subsequence fusion correlation of the factory processing capacity data, and the random residual subsequence fusion correlation of the factory processing capacity data; μ 1,xw ,μ 2,xw ,μ 3,xw They are the periodic subsequence fusion correlation of customer attention data, the trend subsequence fusion correlation of customer attention data, and the random residual subsequence fusion correlation of customer attention data; S23: performing feature selection and feature scaling on the correlation sequence obtained in S22 and inputting the result into the customer relationship-machine learning model, and performing feature compression on the features obtained by feature selection and predicting the potential market value of the processing order of the quartz device; The potential market value of the processing orders for the quartz devices is Φ k for: Among them, α1, α2, and α3 are the value factors of quartz device processing parameter data, factory processing capability data, and customer attention data, respectively.

4. The operation and maintenance method based on machine learning according to claim 3, characterized in that: In S3, the high-dimensional cluster analysis of the potential market value of the processing orders for quartz devices includes the following specific steps: S31: The predicted potential market value of the order is formed into an N-dimensional data feature set L, denoted as L = {a k,n }, where a k,n is the nth characteristic value of the kth quartz device processing order; the data feature set is fuzzy processed by the membership function to obtain the fuzzy processed feature data set H, which is recorded as H = {b k,n }, where b k,n Corresponding to a k,n The characteristic value after fuzzy processing; n represents the number of types of characteristic values; S32: According to S31, extract the fuzzy processed feature data set H, and map the feature data set H to a high-dimensional plane, divide the predicted potential market value of the order into Q categories, and set M cluster center data points, where the cluster center data point set of each potential market value category is G = {g1, g2, ..., g m }, where the density of feature data points on the high-dimensional plane is ρ; g m is the data point of the mth cluster center; Q = M; S33: Construct membership degree d k,m The classification matrix is ​​used to divide and cluster the feature data set H on the high-dimensional plane, where d k,m is the membership degree of the k-th quartz device processing order in the feature data set H to the m-th cluster center data point, where m∈(0,M].

5. An operation and maintenance system based on machine learning, which is used to implement an operation and maintenance method based on machine learning as described in any one of claims 1 to 4, characterized in that: The system includes the following modules: Historical data collection module, sequence prediction module, cluster analysis module, customer classification module, and operation and maintenance decision generation module; The historical data collection module is used to collect historical order data, and perform data preprocessing on the collected historical order data to obtain a processing data sequence and an order data sequence; The sequence prediction module is used to establish a customer relationship-machine learning model, input the processing data sequence and the order data sequence into the customer relationship-machine learning model, and predict the number of processing orders and the total order amount of quartz devices; The cluster analysis module establishes a cluster analysis model according to the number of processing orders and the total order amount of quartz devices, and performs high-dimensional cluster analysis on the number of processing orders and the total order amount of quartz devices; The customer classification module is used to calculate the stability of customers related to different quartz device processing orders according to the cluster analysis results and the stability calculation strategy of related customers, and classify the enterprise customers according to the stability of the customers to obtain the classification data of the enterprise customers; The operation and maintenance decision generating module generates operation and maintenance decision data according to the classification data of the enterprise customers.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, an operation and maintenance method based on machine learning as described in any one of claims 1 to 4 is implemented.

7. An electronic device, characterized in that: include: A memory for storing instructions; A processor is used to execute the instructions so that the device performs the steps of implementing an operation and maintenance method based on machine learning as described in any one of claims 1 to 4.

Images