A matching relationship established rental management system and method

Abstract

The application discloses a renting and purchasing management system and method based on matching relationship establishment, relates to the field of mechanical equipment software leasing management systems, and provides the system which can construct and optimize a two-dimensional mechanical equipment combination matrix, generates a user feature matrix and an article feature matrix through matrix decomposition, and dynamically extracts an equipment feature vector. The system further calculates the similarity between the equipment through an implicit matching coefficient, generates a recommended list of equipment combinations with high matching, and realizes intelligent recommendation of matching leasing.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of mechanical equipment software leasing management system, and particularly relates to a leasing and purchasing management system and method with matched relationship establishment. BACKGROUND

[0002] In the mechanical equipment leasing industry, traditional leasing management systems usually rely on manual or simple online systems to complete the recommendation and matching management of equipment leasing. These systems usually rely on static equipment information and leasing demand data, and are difficult to dynamically capture the specific needs of users, historical leasing data and real-time leasing status of equipment. Traditional leasing management systems generally include a client and a back-end, connected through an API interface to realize the basic data transmission between the client and the back-end. Users can view the availability of equipment on the client and submit leasing requests to the back-end. The back-end then processes the leasing request of the user according to the equipment data in the system. However, this system focuses more on the leasing process of a single device and does not have an intelligent matching recommendation function, making it difficult to effectively solve the combined needs of mechanical equipment. Since mechanical equipment often needs to be used in combination with other equipment in some scenarios, it makes single leasing recommendation difficult to meet the multi-equipment leasing needs in complex application scenarios. For example, construction sites, engineering construction, mineral exploitation and other occasions often involve the use of multiple devices, including excavators, cranes, transport vehicles and other equipment, which need to be recommended in combination to meet actual needs.

[0003] Currently, some more advanced rental management systems have begun to introduce collaborative filtering recommendation algorithms, hoping to provide personalized equipment recommendations for users through user historical data and equipment usage records. However, the recommendation process of these systems often stops at recommending between users and a single device, making it difficult to handle scenarios involving the use of multiple devices in combination. In actual operation, the traditional system has the defect of lacking the ability to integrate and analyze multi-dimensional data. Since the rental demand of equipment is closely related to equipment characteristics, rental frequency, customer historical rental behavior, and other factors, the traditional system often cannot fully utilize these data to achieve accurate recommendations. First, the traditional system lacks data-driven analysis and intelligent recommendation capabilities in terms of equipment combination recommendations, making it difficult to provide effective equipment combination recommendations and affecting user experience. For example, in engineering projects, the rental demand for different equipment often changes according to project progress or construction period requirements, and single-device recommendations cannot meet these changing needs. Second, the traditional system relies on fixed rules or static parameters for the calculation of equipment matching coefficients, lacking dynamic adjustment capabilities. Since equipment may have matching needs in different application scenarios, the traditional system often cannot flexibly adjust the recommendation strategy according to actual usage, resulting in insufficient accuracy of the recommendation results. In addition, the real-time data transmission efficiency of the traditional system is limited, making it difficult to efficiently handle a large number of access requests and return real-time equipment rental data and recommendation results. Especially during peak rental periods, the data processing speed of the traditional system can become a bottleneck, affecting user rental experience. SUMMARY

[0004] The present application proposes a rental and purchase management system and method for establishing matching relationships, which realizes intelligent and dynamic equipment combination recommendations through the introduction of various technical means.

[0005] Among them, a rental and purchase management system for establishing matching relationships includes a client and a backend, the client and the backend are connected through an API interface, the client is used to send access requests to the backend, the backend is used to receive access requests and send real-time rental data to the client, the client sends user rental requests to the backend according to real-time rental data, and the access request includes user rental data;

[0006] Further:

[0007] The system also includes a server, the server is connected to the database of the client and the backend respectively, and extracts the user rental data of the client and the mechanical equipment rental data in the database of the backend respectively, and the server is used for matching rental analysis according to the extracted user rental data and mechanical equipment rental data;

[0008] Among them, the server specifically includes:

[0009] a data extraction unit configured to extract user rental data of the client and mechanical equipment rental data in a database of the background end, and construct a two-dimensional mechanical equipment combination matrix according to the extracted data;

[0010] a feature extraction unit configured to decompose the mechanical equipment combination matrix into a user feature matrix and an item feature matrix, and extract a feature vector of the mechanical equipment in the matrix according to the user feature matrix and the item feature matrix;

[0011] a hidden matching coefficient calculation unit configured to calculate the similarity between the mechanical equipment according to the feature vector of the mechanical equipment, and generate a mechanical equipment matching coefficient matrix;

[0012] a mechanical equipment combination recommendation unit configured to recommend a mechanical equipment combination meeting the user demand according to the hidden matching coefficient, and send mechanical equipment combination recommendation data;

[0013] After the client sends an access request to the background end, the background end and the server simultaneously receive the access request, and the server and the background end respectively send mechanical equipment combination recommendation data and real-time rental data to the client.

[0014] Further, in the data extraction unit, the two-dimensional mechanical equipment combination matrix is used to represent the combination rental situation between mechanical equipment E i and E j , and the mechanical equipment combination matrix is represented as R(i,j), wherein R represents a two-dimensional matrix, W i and E j represent two different mechanical equipment IDs, and i and j represent the indices of the two different mechanical equipment, respectively.

[0015] Further, the two-dimensional matrix R is specifically represented as:

[0016]

[0017] wherein n represents the total number of mechanical equipment.

[0018] Further, in the feature extraction unit, the decomposition of the mechanical equipment combination matrix into the user feature matrix and the item feature matrix is specifically: the mechanical equipment combination matrix is decomposed into the product of two low-rank matrices, that is:

[0019] R≈P·Q T ;

[0020] wherein R represents a two-dimensional matrix, P is a user feature matrix of the mechanical equipment, that is, the latent feature of the mechanical equipment in the user evaluation or rental behavior, Q is an item feature matrix of the mechanical equipment, representing the feature vector of the mechanical equipment in the mechanical equipment combination, and T represents transposition.

[0021] Further, the implicit matching coefficient calculation unit specifically comprises:

[0022] The implicit matching coefficient calculation subunit is configured to calculate the matching coefficient by the similarity of the extracted mechanical equipment feature vectors.

[0023] The initialization matrix subunit is configured to create an initial n×n matrix, which is used to store the matching coefficient between each pair of mechanical equipment.

[0024] The matching coefficient subunit is configured to fill the calculated matching coefficient value into the corresponding position in the matrix.

[0025] The matrix post-processing subunit is configured to normalize the matrix by maximum and minimum scaling.

[0026] Further, in the implicit matching coefficient calculation subunit, the matching coefficient is specifically represented as:

[0027]

[0028] wherein, the S(i,j) represents the matching coefficient of the mechanical equipment E i and E j , the P(i,:) and Q(j,:) respectively represent the feature vectors of the mechanical equipment E i and E j , and the ||P(i,:)|| and ||Q(j,:)|| respectively represent the feature vector lengths of the mechanical equipment E i and E j .

[0029] Further, a matching relationship establishment rental management method is implemented based on the matching relationship establishment rental management system according to any one of the above, and the method is applied to the server and comprises the following steps:

[0030] S1. receiving an access request sent by a client, wherein the access request comprises user rental data;

[0031] S2. extracting the user rental data of the client and the mechanical equipment rental data in the database of the background end;

[0032] S3. performing matching rental analysis according to the extracted user rental data and mechanical equipment rental data;

[0033] wherein, the step S3 specifically comprises the following substeps:

[0034] S301. extracting the user rental data of the client and the mechanical equipment rental data in the database of the background end, and constructing a two-dimensional mechanical equipment combination matrix according to the extracted data;

[0035] S302. Decompose the mechanical equipment combination matrix into a user feature matrix and an item feature matrix, and extract the feature vectors of the mechanical equipment in the matrix according to the user feature matrix and the item feature matrix;

[0036] S303. Calculate the similarity between the mechanical equipment according to the feature vectors of the mechanical equipment, and generate a mechanical equipment matching coefficient matrix;

[0037] S304. Recommend a mechanical equipment combination that meets the user's demand according to the implicit matching coefficient, and send the mechanical equipment combination recommendation data to the client.

[0038] Further, in step S302, the mechanical equipment combination matrix is decomposed into a user feature matrix and an item feature matrix, specifically: the mechanical equipment combination matrix is decomposed into the product of two low-rank matrices, that is:

[0039] R≈P·Q T ;

[0040] Wherein, R represents a two-dimensional matrix, P is the user feature matrix of the mechanical equipment, that is, the latent features of the mechanical equipment in the user evaluation or rental behavior, Q is the item feature matrix of the mechanical equipment, which represents the feature vector of the mechanical equipment in the mechanical equipment combination, and T represents transposition.

[0041] Further, the step S303 specifically includes the following sub-steps:

[0042] S3031. Calculate the matching coefficient by the similarity of the extracted mechanical equipment feature vectors;

[0043] S3032. Create an initial matrix of n x n, which is used to store the matching coefficient between each pair of mechanical equipment;

[0044] S3033. Fill the calculated matching coefficient value into the corresponding position in the matrix;

[0045] S3034. Normalize the matrix by maximum and minimum scaling.

[0046] Further, in step S3031, the calculation of the matching coefficient is specifically represented as:

[0047]

[0048] Wherein, S(i,j) represents the matching coefficient of the mechanical equipment E i and E j , P(i,:) and Q(j,:) represent the mechanical equipment E i and E jcharacteristic vectors of the mechanical equipment E i and E j characteristic vectors of the mechanical equipment E

[0049] The application has the following beneficial effects:

[0050] The system provided by the application can construct and optimize a two-dimensional mechanical equipment combination matrix, and generate a user characteristic matrix and an item characteristic matrix through matrix decomposition, so as to dynamically extract equipment characteristic vectors. The system further calculates the similarity between equipment through implicit matching coefficients, so as to generate a recommended list of equipment combinations with high matching, thereby realizing intelligent recommendation of matching rental. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 The system data flow diagram of the matching relationship established rental management system provided by the embodiment of the application. DETAILED DESCRIPTION

[0052] The technical solutions of the application will be further described in detail below with reference to the accompanying drawings, but the protection scope of the application is not limited to the following description.

[0053] In order to make the purpose, technical solutions and advantages of the application clearer, the application will be further described in detail in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application, that is, the described embodiments are only a part of the embodiments of the application, but not all the embodiments. The components of the embodiments of the application described and shown in the drawings can be arranged and designed in various different configurations.

[0054] Therefore, the detailed description of the embodiments of the application provided in the drawings below is not intended to limit the scope of the claimed application, but only represents selected embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the application. It should be noted that the relationship terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations.

[0055] Also, the term "comprise", "comprising", or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that includes the recited element.

[0056] The features and characteristics of the present application are further described in detail below with reference to the embodiments.

[0057] Wherein, as Figure 1 A rental management system with a matching relationship, comprising a client and a background end, the client and the background end are connected through an API interface, the client is used to send an access request to the background end, the background end is used to receive the access request and send real-time rental data to the client, the client sends a user rental request to the background end according to the real-time rental data, the access request includes user rental data;

[0058] Further, the system further comprises a service end, the service end is connected with the database of the client and the background end respectively, the user rental data of the client and the mechanical equipment rental data in the database of the background end are extracted respectively, the service end is used to carry out matching rental analysis according to the extracted user rental data and mechanical equipment rental data;

[0059] Among them, the service end specifically includes:

[0060] A data extraction unit is used to extract the user rental data of the client and the mechanical equipment rental data in the database of the background end, and construct a two-dimensional mechanical equipment combination matrix according to the extracted data;

[0061] A feature extraction unit is used to decompose the mechanical equipment combination matrix into a user feature matrix and an item feature matrix, and extract the feature vector of the mechanical equipment in the matrix according to the user feature matrix and the item feature matrix;

[0062] An implicit matching coefficient calculation unit is used to calculate the similarity between the mechanical equipment according to the feature vector of the mechanical equipment, and generate a mechanical equipment matching coefficient matrix;

[0063] A mechanical equipment combination recommendation unit is used to recommend a mechanical equipment combination meeting the user's demand according to the implicit matching coefficient, and send mechanical equipment combination recommendation data;

[0064] The client sends an access request to the background end, and the background end and the server receive the access request at the same time, and the server and the background end send mechanical equipment combination recommendation data and real-time rental data to the client respectively.

[0065] Specifically, the user rental data in the client reflects the rental demand, preference and other characteristics of the user; the background end database contains the rental data of all mechanical equipment, recording the type, purpose, rental situation and other information of each device. The data extraction unit obtains the above two types of data through the API interface respectively, and integrates them into a two-dimensional mechanical equipment combination matrix. The elements in the matrix represent the preference degree or actual rental situation of the user for a specific mechanical equipment.

[0066] Further, in the data extraction unit, the two-dimensional mechanical equipment combination matrix is used to represent the combination rental situation between mechanical equipment E i and E j , and the mechanical equipment combination matrix is represented as R(i,j), wherein R represents a two-dimensional matrix, W i and E j represent the ID of two different mechanical equipment respectively, and i and j represent the index of two different mechanical equipment respectively.

[0067] Further, the two-dimensional matrix R is specifically represented as:

[0068]

[0069] Wherein, n represents the total number of mechanical equipment.

[0070] Further, in the feature extraction unit, the mechanical equipment combination matrix is decomposed into a user feature matrix and an item feature matrix, specifically: the mechanical equipment combination matrix is decomposed into the product of two low-rank matrices, that is:

[0071] R≈P·Q T ;

[0072] Wherein, R represents a two-dimensional matrix, P is a user feature matrix of mechanical equipment, that is, the latent feature of mechanical equipment in user evaluation or rental behavior, Q is an item feature matrix of mechanical equipment, representing the feature vector of mechanical equipment in mechanical equipment combination, and T represents transposition. Specifically, the above matrix decomposition can preferably use singular value decomposition (SVD) or non-negative matrix factorization (NMF). Matrix decomposition is used to convert a sparse matrix into a low-rank matrix to extract the latent matching relationship between devices. For example, non-negative matrix factorization is selected:

[0073]

[0074] Wherein, P ikrepresents the value of the ith device on the kth latent feature, and the k represents the number of latent features, ranging from 1 to d, where d is the feature dimension (i.e., the number of latent features) selected at the time of decomposition. jk represents the value of the ith device on the kth latent feature, and the k represents the number of latent features, ranging from 1 to d, where d is the feature dimension (i.e., the number of latent features) selected at the time of decomposition.

[0075] For P: update each row step by step according to the current Q value, so that R≈P·Q T ; for Q, update each column step by step under the condition of fixed P, and optimize the overall reconstruction error.

[0076] The above process is iterated until convergence, i.e., the loss function changes tend to zero or meet the accuracy requirement.

[0077] Further, as a preferred embodiment of the above embodiment, in order to find the optimal p and Q value, it is necessary to define and minimize the loss function, specifically:

[0078] L=∑ (i,j)∈R (R(i,j)-(P·Q T )(i,j)) 2 +λ(||P|| 2 +||Q|| 2 );

[0079] wherein, the R(i,j)-(P·Q T )(o,j)) 2 represents the reconstruction error, used to measure the similarity of matrix R and P·Q T , the λ is a regularization coefficient, and the ||P|| 2 +||Q|| 2 is a regularization term, which avoids overfitting by constraining P and Q values.

[0080] Further, the latent matching coefficient calculation unit specifically comprises:

[0081] The latent matching coefficient calculation subunit is used to calculate the matching coefficient through the similarity of the extracted mechanical equipment feature vectors; specifically, the main task of the latent matching coefficient calculation subunit is to calculate the similarity between each pair of devices to generate the matching coefficient using the feature vectors of the mechanical equipment. After the latent matching coefficient calculation subunit extracts the feature vectors of the mechanical equipment, cosine similarity, Pearson correlation coefficient and other methods are used to evaluate the similarity of each pair of devices, so as to calculate the matching coefficient between them. These matching coefficients are used to represent the degree of cooperation or adaptability between devices, thereby providing a basis for the recommendation unit.

[0082] The initialization matrix subunit is configured to create an initial matrix of n x n, which is used to store the matching coefficients between each pair of mechanical devices. Specifically, the initialization matrix subunit is configured to create an initial matrix for storing the matching coefficients. In the system initialization stage, the subunit generates an empty matrix structure, and the rows and columns of the matrix represent the respective devices in the system. The element values in the matrix are empty or default values at the beginning. The initialized matrix provides storage space for subsequent steps and fills in the calculated matching coefficients of each pair of devices.

[0083] The matching coefficient subunit is configured to fill the calculated matching coefficient values into the corresponding positions in the matrix. The matrix post-processing subunit is configured to normalize the matrix through maximum and minimum scaling. Specifically, the matching coefficient subunit is responsible for filling the calculated matching coefficient values into the corresponding positions in the matrix. For the matching relationship of each pair of devices, once the matching coefficient is calculated, the subunit stores the matching coefficient in the corresponding position of the matching coefficient matrix, and after filling the matrix one by one, the matching coefficient matrix records the matching relationship between all devices.

[0084] Further, in the calculation of the implicit matching coefficient subunit, the matching coefficient is specifically represented as:

[0085]

[0086] wherein S(i, j) represents the matching coefficient of the mechanical devices E i and E j , P(i, :) and Q(j, :) represent the feature vectors of the mechanical devices E i and E j , respectively, and ||P(i, :)|| and ||Q(j, :)|| represent the feature vector lengths of the mechanical devices E i and E j , respectively.

[0087] Specifically, the implementation of device recommendation through the implicit matching coefficient includes two implementation modes, which are:

[0088] Based on the matching coefficient recommendation, for each device, the device with the highest matching coefficient is found from the set of matching coefficients of the device with other devices, and the device combination is recommended.

[0089] A recommendation threshold is set for the matching coefficient, and an exemplary device combination is recommended only when S(i, j) > 0.5.

[0090] Further, P(i, :) is the feature vector of the device E i in the implicit feature space, representing the performance of the device in the "user demand", and Q(j, :) is the feature vector of the device E jThe feature vector in the implicit feature space represents the performance of the device as a "matching item" feature, and the dot product P(i,:)·Q(j,:) measures the directional similarity of the two feature vectors. The modulus normalization in the denominator ensures that the similarity is within the range [0, 1]; in addition, through the above calculation, S(i,j) is obtained, which is the matching coefficient of the device E i and E j .

[0091] The above implementation utilizes the user-item based matrix decomposition technology in collaborative filtering to represent the implicit demand and features of the device as a vectorized feature, thereby performing similarity measurement on the device in the implicit feature space. For example, the performance of some devices in the implicit feature space may be more similar to another type of device, so that matching recommendations can be made through feature similarity.

[0092] Further, a matching relationship establishing rental management method, which is implemented based on any one of the above-mentioned matching relationship establishing rental management systems, is applied to a server and includes the following steps:

[0093] S1. receiving an access request sent by a client, wherein the access request includes user rental data;

[0094] S2. extracting the user rental data of the client and the mechanical equipment rental data in the database of the background end;

[0095] S3. performing matching rental analysis based on the extracted user rental data and mechanical equipment rental data;

[0096] The step S3 specifically includes the following sub-steps:

[0097] S301. extracting the user rental data of the client and the mechanical equipment rental data in the database of the background end, and constructing a two-dimensional mechanical equipment combination matrix based on the extracted data;

[0098] S302. decomposing the mechanical equipment combination matrix into a user feature matrix and an item feature matrix, and extracting feature vectors of the mechanical equipment in the matrix based on the user feature matrix and the item feature matrix;

[0099] S303. calculating the similarity between the mechanical equipment based on the feature vectors of the mechanical equipment, and generating a mechanical equipment matching coefficient matrix;

[0100] S304. recommending a mechanical equipment combination that meets the user's demand based on the implicit matching coefficient, and sending the mechanical equipment combination recommendation data to the client.

[0101] Further, the step S302, the mechanical equipment combination matrix is decomposed into user feature matrix and article feature matrix, specifically: the mechanical equipment combination matrix is decomposed into the product of two low rank matrices, namely:

[0102] R≈P·Q T ;

[0103] Wherein, the R represents two-dimensional matrix, the P is the user feature matrix of mechanical equipment, namely the potential feature of mechanical equipment on the user evaluation or rental behavior, the Q is the article feature matrix of mechanical equipment, and the characteristic vector of mechanical equipment on mechanical equipment combination is represented, and the T represents transposition.

[0104] Further, the step S303 specifically includes the following sub-steps:

[0105] S3031. The matching coefficient is calculated by the similarity of extracted mechanical equipment characteristic vector;

[0106] S3032. The initial matrix of n*n is created, and the initial matrix is used to store the matching coefficient between each pair of mechanical equipment;

[0107] S3033. The matching coefficient value calculated is filled in the corresponding position of the matrix;

[0108] S3034. The matrix is normalized by maximum minimum scaling.

[0109] Further, the step S3031, the matching coefficient is calculated specifically as follows:

[0110]

[0111] Wherein, the S (i, j) represents the matching coefficient of mechanical equipment E i And E j , the P (i, :) and Q (j, :) respectively represent the characteristic vector of mechanical equipment E i And E j , and the ||P (i, :) || and ||Q (j, :) || respectively represent the characteristic vector module length of mechanical equipment E i And E j .

[0112] The above only is the preferred embodiment of the present application, it should be understood that the present application is not limited to the form disclosed herein, should not be regarded as excluding other embodiments, and can be used in various other combinations, modifications and environment, and can be changed within the scope of the concept described herein, by the above teaching or related art or knowledge of personnel, and the change and variation of personnel without departing from the spirit and scope of the present application, then all should be within the protection scope of the claims of the present application.

Claims

1. A matching relationship established rental management system, comprising a client and a background end, the client and the background end are connected through an API interface, the client is used to send an access request to the background end, the background end is used to receive the access request and send real-time rental data to the client, the client sends a user rental request to the background end according to the real-time rental data, and the access request includes user rental data; Characterized in that: The system further comprises a service end, the service end is connected with the database of the client and the background end respectively, the user rental data of the client and the mechanical equipment rental data in the database of the background end are extracted respectively, and the service end is used for matching rental analysis according to the extracted user rental data and the mechanical equipment rental data; Wherein, the service end specifically comprises: A data extraction unit is used for extracting the user rental data of the client and the mechanical equipment rental data in the database of the background end, and constructing a two-dimensional mechanical equipment combination matrix according to the extracted data; A feature extraction unit is used for decomposing the mechanical equipment combination matrix into a user feature matrix and an item feature matrix, and extracting the feature vector of the mechanical equipment in the matrix according to the user feature matrix and the item feature matrix; An implicit matching coefficient calculation unit is used for calculating the similarity between the mechanical equipment according to the feature vector of the mechanical equipment, and generating a mechanical equipment matching coefficient matrix; A mechanical equipment combination recommendation unit is used for recommending a mechanical equipment combination meeting the user demand according to the implicit matching coefficient, and sending mechanical equipment combination recommendation data; After the client sends an access request to the background end, the background end and the service end receive the access request at the same time, and the service end and the background end send mechanical equipment combination recommendation data and real-time rental data to the client respectively; The data extraction unit, a two-dimensional mechanical equipment combination matrix is used to represent the combination leasing situation between the mechanical equipment and The mechanical equipment combination matrix is represented as , wherein the represents a two-dimensional matrix, the and respectively represent two different mechanical equipment IDs, and the and respectively represent the indexes of the two different mechanical equipment. The two-dimensional matrix Specifically represented as: ； Wherein, n represents the total number of mechanical equipment; The decomposition of the mechanical equipment combination matrix into a user feature matrix and an item feature matrix is specifically that the mechanical equipment combination matrix is decomposed into the product of two low-rank matrices, that is: ； wherein the denotes a two-dimensional matrix, the is a user feature matrix of the mechanical device, i.e. potential features of the mechanical device on the user evaluation or rental behavior, the is an item feature matrix of the mechanical device, representing a feature vector of the mechanical device on the mechanical device portfolio, the denotes a transposition.

2. The commitment establishing rental-purchase management system of claim 1, wherein, The implicit matching coefficient calculation unit specifically comprises: A calculation implicit matching coefficient subunit is used for calculating the matching coefficient by the similarity of the extracted mechanical equipment feature vector; initializing a matrix subunit for creating an initial matrix for storing a matching coefficient between each pair of mechanical devices ; A matching coefficient subunit is used for filling the calculated matching coefficient value into the corresponding position in the matrix; Matrix post-processing is used for normalizing the matrix by maximum and minimum scaling.

3. The commitment establishing rental-purchase management system of claim 2, wherein, In the calculation implicit matching coefficient subunit, the calculation of the matching coefficient is specifically represented as: ； Wherein, the indicates a matching coefficient of the mechanical equipment and , the and respectively indicate a feature vector of the mechanical equipment and , the and respectively indicate a feature vector module length of the mechanical equipment and .

4. A method for establishing a rental-purchase management system according to any one of claims 1 to 3, characterized in that, Applied to the service end, comprising the following steps: S1. Receive the access request sent by the client, the access request includes user rental data; S2. Extract the user rental data of the client and the mechanical equipment rental data in the database of the background end; S3. Perform matching rental analysis according to the extracted user rental data and the mechanical equipment rental data; Wherein, the step S3 specifically comprises the following substeps: S301. Extract the user rental data of the client and the mechanical equipment rental data in the database of the background end, and construct a two-dimensional mechanical equipment combination matrix according to the extracted data; S302. Decompose the mechanical equipment combination matrix into a user feature matrix and an item feature matrix, and extract the feature vector of the mechanical equipment in the matrix according to the user feature matrix and the item feature matrix; S303. Calculate the similarity between the mechanical equipment according to the feature vector of the mechanical equipment, and generate a mechanical equipment matching coefficient matrix; S304. According to the implicit matching coefficient, recommend a mechanical equipment combination that meets the user's demand, and send the mechanical equipment combination recommendation data to the client; In the step S302, the mechanical equipment combination matrix is decomposed into a user feature matrix and an item feature matrix, which is specifically: the mechanical equipment combination matrix is decomposed into the product of two low-rank matrices, that is: ； wherein said represents a two-dimensional matrix, said is a user feature matrix of the mechanical equipment, i.e. potential features of the mechanical equipment on the user evaluation or rental behavior, said is an item feature matrix of the mechanical equipment, representing a feature vector of the mechanical equipment on the mechanical equipment combination, said represents a transpose.

5. The method for establishing a package relationship according to claim 4, wherein, The step S303 specifically includes the following sub-steps: S3031. Calculate the matching coefficient through the similarity of the extracted mechanical equipment feature vector; S3032. Create an initial matrix for storing the matching coefficients between each pair of mechanical devices. S3033. Create a matrix for storing the matching coefficients between each pair of mechanical devices. S3033. Fill the calculated matching coefficient value into the corresponding position in the matrix; S3034. Normalize the matrix by maximum and minimum scaling.

6. The method for establishing a package relationship according to claim 5, wherein, In the step S3031, the calculation of the matching coefficient is specifically represented as: ； Wherein, the represents the matching coefficient of mechanical equipment and , the and respectively represent the characteristic vectors of mechanical equipment and , the and respectively represent the characteristic vector lengths of mechanical equipment and .

Images