Inventory data accuracy verification system and method applied to financial cloud secondary development

Abstract

The invention relates to the technical field of data verification, and discloses an inventory data accuracy verification system and method applied to financial cloud secondary development. The method comprises the following steps: firstly, performing data missing filling based on data correlation to obtain a processed inventory data set and a processed ex-warehouse data set; secondly, obtaining an inventory suspected data set through data comparison, and performing preliminary verification on inventory inaccurate data in the inventory suspected data set by using an improved clustering algorithm; optimizing the initial clustering center of the improved clustering detection algorithm by using a multi-strategy fusion emerald optimization algorithm to form a new clustering cluster; and finally, inaccurate data final verification is carried out according to the new clustering cluster, inaccurate inventory data is output, and accuracy verification is completed. By processing and analyzing the inventory data, the purpose of checking the accuracy of the inventory data is achieved, and the method is objective and accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of data verification, and in particular to an inventory data accuracy verification system and method applied to secondary development of financial cloud. Background Art

[0002] Chinese patent CN116662375B discloses a prescription data verification method and system based on HIS, the method specifically includes: obtaining the prescription log in the HIS database, obtaining the medical prescription to be verified information log, collecting the data in the medical prescription to be verified information log, recording it as the medical prescription to be verified information data, and then performing noise reduction processing to obtain the medical prescription to be verified noise reduction data; using the data verification algorithm to verify the medical prescription to be verified noise reduction data, obtain the medical prescription drug verification result, and obtain the drug information data to be added, and add it to the medical information drug update library; obtain the patient information in the HIS database, obtain the patient information data, simulate the order in the medical information drug update library, and generate the medical simulation prescription monitoring result; perform abnormal detection processing on the medical simulation prescription monitoring result, obtain the medical prescription drug accurate verification result, and realize prescription data verification. The data verification method of this invention is cumbersome and time-consuming.

[0003] The traditional data accuracy verification method directly detects abnormal fluctuations in the data to be verified, which is time-consuming and cannot be comprehensive. At the same time, due to the lack of the use of technologies such as artificial intelligence, when verifying the accuracy of inventory data, the inventory data is usually redundant and contains omissions and anomalies, which brings trouble to accuracy verification and is not conducive to helping track the source of data. Summary of the invention

[0004] In view of the problems in the related technology, the present invention provides an inventory data accuracy verification system and method applied to the secondary development of financial cloud to overcome the above-mentioned technical problems existing in the existing related technology.

[0005] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention is a method for verifying the accuracy of inventory data applied to secondary development of financial cloud, comprising the following steps:

[0007] S1. Obtain inventory-related data and outbound delivery-related data in the financial cloud to obtain an initial inventory data set and an initial outbound delivery data set, and then fill in missing data based on data correlation to obtain a processed inventory data set and a processed outbound delivery data set;

[0008] S2. Obtaining a suspected inventory data set based on the processed inventory data set and the processed outbound data set, and performing a preliminary check on inaccurate inventory data in the suspected inventory data set using an improved clustering algorithm in combination with spatial distance and time series to obtain a clustering objective function;

[0009] S3, taking the clustering objective function as the fitness function, using the multi-strategy fusion Pied Kingfisher optimization algorithm to optimize the initial clustering center of the improved clustering detection algorithm, obtaining the optimized clustering center, and generating a new cluster;

[0010] S4. Perform a final check on the inaccurate data based on the new cluster, output the inaccurate inventory data, divide the suspected inventory data set, and complete the accuracy check.

[0011] The invention obtains inventory-related data and delivery-related data, and based on data correlation, fills in missing data for inventory-related data and delivery-related data respectively; this method overcomes the defects of traditional data filling methods, such as long processing time and filling only through prediction, by quantifying the association between data, greatly improving the integrity and accuracy of data, and facilitating subsequent data analysis and processing; secondly, it identifies suspected inventory data, and based on spatial distance and time series, uses an improved clustering algorithm to perform preliminary verification of inaccurate inventory data in the identified suspected inventory data to obtain a clustering objective function; this method clusters the time series and updates the cluster center according to the spatial distance, which is more accurate than the traditional clustering algorithm. The method can effectively identify abnormal time series fluctuations, has more advantages in processing time series data, and has better clustering effects; the clustering objective function is then used as the fitness function, and the initial clustering center is optimized using the Pied Kingfisher optimization algorithm with multi-strategy fusion; the algorithm achieves optimization by simulating the hunting behavior and symbiotic relationship of the Pied Kingfisher. The optimized algorithm uses chaotic mapping to overcome the problem of random initialization unevenness, and the triangular walk strategy increases the local optimization ability of the algorithm. The dynamic probability balance weight is used to better achieve optimization, which greatly reduces the number of iterations of the clustering algorithm and saves a lot of verification time; finally, a final verification of inaccurate data is performed, and the inaccurate inventory data is output to complete the accuracy verification.

[0012] Preferably, the S1 comprises the following steps:

[0013] S11. Obtain inventory-related data in the financial cloud, the inventory-related data including the existing inventory quantity, entry time, inventory value, etc., obtain inventory data, and use the entry time as the entry time point to establish an inventory time series to form an initial inventory data set; then obtain outbound data, the outbound data including the outbound quantity, outbound time, outbound destination, etc., obtain outbound data, and use the outbound time as the outbound time point to establish an outbound time series to form an initial outbound data set;

[0014] S12: fill in missing data according to the data correlation between the initial inventory data set and the initial outbound data set to obtain a processed inventory data set and a processed outbound data set. The specific steps are as follows:

[0015] S121, record the inventory time series in the initial inventory data set as A 1 ={a 1 ,a 2 ,a 3 ,...,a m}, where a m Indicates the mth storage time point, set Indicates The storage time point, Indicates The storage time point is calculated The storage time point and The covariance of the inventory data corresponding to the warehousing time points is recorded as the inventory data covariance, and then the average value of the inventory time series is calculated to calculate the data correlation of the inventory data. The formula is as follows:

[0016]

[0017] Among them, β 1 Represents the data dependencies of inventory data, represents the covariance of inventory data, represents the average value of the inventory time series;

[0018] The outbound time series in the initial outbound data set is recorded as A 2 ={a 1 ′,a′ 2 ,a 3 ′,...,a′ m′}, where a m′ Indicates the m′th outbound time point, set Indicates The time of shipment, Indicates The outbound time point is calculated The outbound time point and The covariance of the outbound data corresponding to the outbound time point is recorded as the outbound data covariance, and then the average value of the outbound time series is calculated to calculate the data correlation of the outbound data. The formula is as follows:

[0019]

[0020] Among them, β 2 Indicates the data relevance of outbound data. represents the covariance of the outbound data, Represents the average value of the outbound time series;

[0021] S122, calculate the The storage time point and The Euclidean distance of the inventory data corresponding to the first entry time point is calculated, and the inventory data corresponding to all entry time points in the inventory time series are accumulated, and then multiplied by the data correlation of the inventory data to obtain the inventory correlation distance; calculate the The outbound time point and The Euclidean distance of the outbound data corresponding to the outbound time points is calculated, and the outbound data corresponding to all the outbound time points in the outbound time series are accumulated, and then multiplied by the data correlation of the outbound data to obtain the outbound correlation distance;

[0022] For missing inventory data in the initial inventory data set Setting missing inventory data There exists k 1 neighboring inventory data, calculate the inventory related distance and k 1 The average value of the product of adjacent inventory data is obtained to obtain the inventory data filling value, and the missing inventory data is filled with the inventory data. Fill in missing data, fill in all missing inventory data in the initial inventory data set in turn, and obtain a processed inventory data set; fill in missing outbound data in the initial outbound data set Set missing outbound data There exists k 2 neighboring outbound data, calculate the outbound related distance and k 2 The average value of the product of the adjacent outbound data is obtained to obtain the outbound data filling value, and the outbound data filling value is used to fill the missing outbound data. Perform data missing filling, fill all missing outbound data in the initial outbound data set in turn, and obtain a processed outbound data set.

[0023] The invention obtains inventory-related data and delivery-related data, and based on data correlation, fills in missing data for inventory-related data and delivery-related data respectively. By quantifying the correlation between the data, the invention overcomes the defects of traditional data filling methods such as long processing time and filling only through prediction, greatly improves the integrity and accuracy of the data, and facilitates subsequent data analysis and processing.

[0024] Preferably, S2 comprises the following steps:

[0025] S21, obtaining the total data of inbound and outbound, obtaining an inbound and outbound data set, and then accumulating the corresponding data in the processed inventory data set and the processed outbound data set according to the inventory time series and the outbound time series to obtain a data set to be verified; setting a verification threshold, comparing the data set to be verified with the inbound and outbound data set, when the difference between the corresponding data in the data set to be verified and the inbound and outbound data set is greater than the verification threshold, at this time, the corresponding data in the data set to be verified is recorded as suspected inventory data, otherwise it is recorded as normal inventory data, and the inventory time point is recorded to obtain a suspected inventory data set;

[0026] S22. Combining spatial distance and time series, using an improved clustering algorithm to perform preliminary verification on the inaccurate inventory data in the suspected inventory data set to obtain a clustering objective function. The specific steps are as follows:

[0027] S221, selecting any inventory time point in the inventory suspected data set as the initial cluster center, constructing a topological space, calculating the spatial distance between the initial cluster center and other inventory time points in the topological space, and obtaining a new cluster center. The specific steps are as follows:

[0028] S2211. All inventory time points in the inventory suspected data set are recorded as suspected time series and mapped to the topological space. The origin coordinates in the topological space are marked as B(0,0,0). Each spatial node in the topological space is set to represent an inventory time point to obtain the coordinates of the inventory time point. Taking the initial cluster center as the starting point, the spatial distance from the initial cluster center to other inventory time points is calculated to obtain the closest spatial distance. The inventory time point corresponding to the closest spatial distance is recorded as the closest spatial distance inventory time point.

[0029] S2212, recording the closest spatial distance inventory time point as a new cluster center, and using the new cluster center to replace the initial cluster center;

[0030] S222, continue to calculate the spatial distance between the new cluster center and the nearest inventory time point, generate an updated cluster center, and then calculate the spatial distance between the updated cluster center and other inventory time points to obtain the inventory time point farthest from the updated cluster center, record the suspected inventory data corresponding to the farthest inventory time point as inaccurate data, and complete the preliminary verification of inaccurate data; use the clustering objective function to represent the spatial distance between the new cluster center and the updated cluster center, and the calculation formula is as follows:

[0031]

[0032] Among them, F represents the clustering objective function, χ represents the suspected time series, d represents the new cluster center, c iRepresents the i-th inventory time point in the suspected time series.

[0033] This invention identifies suspected inventory data, clusters the time series based on spatial distance and time series, updates the cluster centers according to the spatial distance, and uses an improved clustering algorithm to perform preliminary verification of inaccurate inventory data in the identified suspected inventory data. Compared with traditional clustering algorithms, it can effectively identify abnormal time series fluctuations, has more advantages in processing time series data, and has better clustering effects.

[0034] Preferably, S3 comprises the following steps:

[0035] S31, taking the clustering objective function as the fitness function, introducing chaos mapping, triangle walk strategy and dynamic probability switching strategy to improve the piebald kingfisher optimization algorithm, obtaining a multi-strategy fusion piebald kingfisher optimization algorithm, using the multi-strategy fusion piebald kingfisher optimization algorithm to optimize the initial cluster center of the improved cluster detection algorithm, and obtaining the optimized cluster center, the specific steps are as follows:

[0036] S311, setting a search space with dimension j, in which there is a population of spotted kingfishers, taking the clustering objective function as the fitness function, the initial cluster center update process is the continuous iteration process of the spotted kingfisher population, taking the minimum clustering objective function value as the optimal fitness function value; using chaotic mapping to initialize the spotted kingfisher population, setting the current number of iterations to t, and the position of the e-th spotted kingfisher individual in the t-th iteration spotted kingfisher population to C e (t), the position of the g-th individual of the Pied Kingfisher population in the t-th iteration is C g (t), d 1 represents a random number between the interval [0, 1], the chaos parameter is φ, the number of chaotic sequence particles is c, when C e When (t)<0.5, the position of the e-th Pied Kingfisher in the t+1th iteration When C e When (t)≥0.5, the position of the e-th Pied Kingfisher in the t+1th iteration

[0037] The Pied Kingfisher population enters the exploration phase, which includes the perching strategy and the hovering strategy. The normally distributed random number between the interval [1, j] is set as d 2 , the maximum number of iterations is T,d 3 Represents a random number between the interval [0, 1], habitat parameter The position C of the e-th individual of the Pied Kingfisher in the t+1th iteration of the roosting strategy e (t+1)=C e (t)+(2d 2 -1)·γ1 ·(C g (t)-C e (t)); calculate the fitness function values ​​of the e-th and g-th individual of the spotted kingfisher population in the t-th iteration, denoted as f e and f g , hover parameter The position C of the e-th individual of the Pied Kingfisher in the t+1th iteration when using the circling strategy e (t+1)=C e (t)+(2d 2 -1)·γ 2 ·(C g (t)-C e (t)); Use dynamic probability switching strategy to balance the proportion of perching strategy and hovering strategy, dynamic probability By comparing the dynamic probability and the random number d 3 Dynamically switch between perching strategy and hovering strategy;

[0038] S312, at this time, calculate the fitness function value corresponding to the individual spotted kingfisher in the spotted kingfisher population, find the current best fitness function value as the current optimal clustering center, and continue to optimize the current optimal clustering center; the spotted kingfisher population enters the development stage, the spotted kingfisher population begins to prey, introduces the triangle wandering strategy to update the position of the spotted kingfisher individual, and calculates the distance l between the e-th spotted kingfisher and the g-th spotted kingfisher individual in the t-th iteration of the spotted kingfisher population 1 , set d 4 Represents a random number between the interval [0, 1], with a walking step length of l 2 = l 1 ·d 4 , walk parameter η = l 1 2 +l 2 2 -2l 1 ·l 2 ·cos(2πd 4 ), for position C e (t+1) is updated, at this time C e (t+1)=C e (t)+η·d 4 ; The Pied Kingfisher population enters the escape stage. Calculate the unpredated efficiency of the Pied Kingfisher population at this time, and set d 5 Represents a random number between the interval [0, 1]. When the random number d 5 When it is greater than the unpredated efficiency, the position C is updated again according to the hunting ability of the individual kingfisher e(t+1), otherwise the position is not updated, and the final position of the individual of the spotted kingfisher in the t+1th iteration is obtained, and the next generation of the spotted kingfisher population is generated and iterated until the current number of iterations reaches the maximum number of iterations, then the iteration is stopped, and the final position of the spotted kingfisher is obtained. The individual position of the spotted kingfisher corresponding to the best fitness function value is recorded as the optimized cluster center;

[0039] S32, replacing the initial cluster center with the optimized cluster center, clustering the inventory time points according to the spatial distance, continuously updating the initial cluster center until the position of the cluster center no longer changes, obtaining the final cluster center, forming a cluster with the final cluster center, and obtaining a new cluster.

[0040] This invention optimizes the initial clustering centers by using a multi-strategy fusion kingfisher optimization algorithm, and achieves optimization by simulating the hunting behavior and symbiotic relationship of the kingfisher. The optimized algorithm uses chaotic mapping to overcome the problem of random uneven initialization, and the triangular walk strategy increases the local optimization ability of the algorithm. The dynamic probability balance weight is used to better achieve optimization, which greatly reduces the number of iterations of the clustering algorithm and saves a lot of verification time.

[0041] Preferably, the S4 comprises the following steps:

[0042] S41. Set the number of new clusters to h, the new clusters include several inventory time points, set a distance threshold, calculate the distance between the final cluster center and the inventory time point in the new cluster, when the distance between the final cluster center and the inventory time point is greater than the distance threshold, record the corresponding inventory time point as an inaccurate inventory time point, identify the suspected inventory data corresponding to the inaccurate inventory time point in the suspected inventory data set, obtain inaccurate inventory data, divide the suspected inventory data set into an inaccurate inventory data set and an accurate inventory data set, and complete the inventory data accuracy verification.

[0043] This embodiment also discloses a system for inventory data accuracy verification method applied to financial cloud secondary development, which specifically includes: a data missing filling module, an inaccurate data preliminary verification module, a cluster center optimization module and an inaccurate data final verification module;

[0044] The data missing filling module is used to fill the data missing of the inventory data set based on data relevance;

[0045] The inaccurate data preliminary verification module is used to perform preliminary verification on inaccurate data in the suspected inventory data using an improved clustering algorithm;

[0046] The cluster center optimization module is used to optimize the initial cluster center using a multi-strategy fusion Kingfisher optimization algorithm;

[0047] The inaccurate data final verification module is used to identify inaccurate inventory data in the inventory suspected data set and complete accuracy verification.

[0048] The present invention has the following beneficial effects:

[0049] 1. The present invention obtains inventory-related data and delivery-related data, and based on data correlation, fills in missing data for inventory-related data and delivery-related data respectively. By quantifying the correlation between the data, the present invention overcomes the defects of traditional data filling methods such as long processing time and filling through prediction alone, greatly improves the integrity and accuracy of the data, and facilitates subsequent data analysis and processing.

[0050] 2. The invention identifies suspected inventory data, clusters the time series based on spatial distance and time series, updates the cluster center according to the spatial distance, and uses an improved clustering algorithm to perform preliminary verification of inaccurate inventory data in the identified suspected inventory data. Compared with traditional clustering algorithms, it can effectively identify abnormal time series fluctuations, has more advantages in processing time series data, and has better clustering effects.

[0051] 3. The invention optimizes the initial clustering centers by using a multi-strategy fusion kingfisher optimization algorithm, and realizes optimization by simulating the hunting behavior and symbiotic relationship of the kingfisher. The optimized algorithm uses chaotic mapping to overcome the problem of random uneven initialization, and the triangular walk strategy increases the local optimization ability of the algorithm. The dynamic probability balance weight is used to better realize optimization, which greatly reduces the number of iterations of the clustering algorithm and saves a lot of verification time.

[0052] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying any creative work.

[0054] Figure 1 The present invention provides a flow chart of inventory data accuracy verification in an inventory data accuracy verification system applied to secondary development of a financial cloud. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0057] Example 1

[0058] See also Figure 1 This implementation discloses an inventory data accuracy verification method applied to the secondary development of financial cloud, which specifically includes the following contents:

[0059] S1. Obtain inventory-related data and outbound delivery-related data in the financial cloud to obtain an initial inventory data set and an initial outbound delivery data set, and then fill in missing data based on data correlation to obtain a processed inventory data set and a processed outbound delivery data set;

[0060] The S1 comprises the following steps:

[0061] S11. Obtain inventory-related data in the financial cloud, the inventory-related data including the existing inventory quantity, entry time, inventory value, etc., obtain inventory data, and use the entry time as the entry time point to establish an inventory time series to form an initial inventory data set; then obtain outbound data, the outbound data including the outbound quantity, outbound time, outbound destination, etc., obtain outbound data, and use the outbound time as the outbound time point to establish an outbound time series to form an initial outbound data set;

[0062] S12: fill in missing data according to the data correlation between the initial inventory data set and the initial outbound data set to obtain a processed inventory data set and a processed outbound data set. The specific steps are as follows:

[0063] S121, record the inventory time series in the initial inventory data set as A 1 ={a 1 ,a 2 ,a 3 ,...,a m}, where a m Indicates the mth storage time point, set Indicates The storage time point, Indicates The storage time point is calculated The storage time point and The covariance of the inventory data corresponding to the warehousing time points is recorded as the inventory data covariance, and then the average value of the inventory time series is calculated to calculate the data correlation of the inventory data. The formula is as follows:

[0064]

[0065] Among them, β 1 Represents the data dependencies of inventory data, represents the covariance of inventory data, represents the average value of the inventory time series;

[0066] The outbound time series in the initial outbound data set is recorded as A 2 ={a′ 1 ,a′ 2 ,a′ 3 ,...,a′ m′}, where a m′ Indicates the m′th outbound time point, set Indicates The time of shipment, Indicates The outbound time point is calculated The outbound time point and The covariance of the outbound data corresponding to the outbound time point is recorded as the outbound data covariance, and then the average value of the outbound time series is calculated to calculate the data correlation of the outbound data. The formula is as follows:

[0067]

[0068] Among them, β 2 Indicates the data relevance of outbound data. represents the covariance of the outbound data, Represents the average value of the outbound time series;

[0069] S122, calculate the The storage time point and The Euclidean distance of the inventory data corresponding to the first entry time point is calculated, and the inventory data corresponding to all entry time points in the inventory time series are accumulated, and then multiplied by the data correlation of the inventory data to obtain the inventory correlation distance; calculate the The outbound time point and The Euclidean distance of the outbound data corresponding to the outbound time points is calculated, and the outbound data corresponding to all the outbound time points in the outbound time series are accumulated, and then multiplied by the data correlation of the outbound data to obtain the outbound correlation distance;

[0070] For missing inventory data in the initial inventory data set Setting missing inventory data There exists k 1 neighboring inventory data, calculate the inventory related distance and k 1 The average value of the product of adjacent inventory data is obtained to obtain the inventory data filling value, and the missing inventory data is filled with the inventory data. Fill in missing data, fill in all missing inventory data in the initial inventory data set in turn, and obtain a processed inventory data set; fill in missing outbound data in the initial outbound data set Set missing outbound data There exists k 2 neighboring outbound data, calculate the outbound related distance and k 2 The average value of the product of the adjacent outbound data is obtained to obtain the outbound data filling value, and the outbound data filling value is used to fill the missing outbound data. Fill in missing data, fill in all missing outbound data in the initial outbound data set in turn, and obtain a processed outbound data set;

[0071] S2. Obtaining a suspected inventory data set based on the processed inventory data set and the processed outbound data set, and performing a preliminary check on inaccurate inventory data in the suspected inventory data set using an improved clustering algorithm in combination with spatial distance and time series to obtain a clustering objective function;

[0072] The S2 comprises the following steps:

[0073] S21, obtaining the total data of inbound and outbound, obtaining an inbound and outbound data set, and then accumulating the corresponding data in the processed inventory data set and the processed outbound data set according to the inventory time series and the outbound time series to obtain a data set to be verified; setting a verification threshold, comparing the data set to be verified with the inbound and outbound data set, when the difference between the corresponding data in the data set to be verified and the inbound and outbound data set is greater than the verification threshold, at this time, the corresponding data in the data set to be verified is recorded as suspected inventory data, otherwise it is recorded as normal inventory data, and the inventory time point is recorded to obtain a suspected inventory data set;

[0074] S22. Combining spatial distance and time series, using an improved clustering algorithm to perform preliminary verification on the inaccurate inventory data in the suspected inventory data set to obtain a clustering objective function. The specific steps are as follows:

[0075] S221, selecting any inventory time point in the inventory suspected data set as the initial cluster center, constructing a topological space, calculating the spatial distance between the initial cluster center and other inventory time points in the topological space, and obtaining a new cluster center. The specific steps are as follows:

[0076] S2211. All inventory time points in the inventory suspected data set are recorded as suspected time series and mapped to the topological space. The origin coordinates in the topological space are marked as B(0,0,0). Each spatial node in the topological space is set to represent an inventory time point to obtain the coordinates of the inventory time point. Taking the initial cluster center as the starting point, the spatial distance from the initial cluster center to other inventory time points is calculated to obtain the closest spatial distance. The inventory time point corresponding to the closest spatial distance is recorded as the closest spatial distance inventory time point.

[0077] S2212, recording the closest spatial distance inventory time point as a new cluster center, and using the new cluster center to replace the initial cluster center;

[0078] S222, continue to calculate the spatial distance between the new cluster center and the nearest inventory time point, generate an updated cluster center, and then calculate the spatial distance between the updated cluster center and other inventory time points to obtain the inventory time point farthest from the updated cluster center, record the suspected inventory data corresponding to the farthest inventory time point as inaccurate data, and complete the preliminary verification of inaccurate data; use the clustering objective function to represent the spatial distance between the new cluster center and the updated cluster center, and the calculation formula is as follows:

[0079]

[0080] Among them, F represents the clustering objective function, χ represents the suspected time series, d represents the new cluster center, c i represents the i-th inventory time point in the suspected time series;

[0081] S3, taking the clustering objective function as the fitness function, using the multi-strategy fusion Pied Kingfisher optimization algorithm to optimize the initial clustering center of the improved clustering detection algorithm, obtaining the optimized clustering center, and generating a new cluster;

[0082] The S3 comprises the following steps:

[0083] S31, taking the clustering objective function as the fitness function, introducing chaos mapping, triangle walk strategy and dynamic probability switching strategy to improve the piebald kingfisher optimization algorithm, obtaining a multi-strategy fusion piebald kingfisher optimization algorithm, using the multi-strategy fusion piebald kingfisher optimization algorithm to optimize the initial cluster center of the improved cluster detection algorithm, and obtaining the optimized cluster center, the specific steps are as follows:

[0084] S311, setting a search space with dimension j, in which there is a population of spotted kingfishers, taking the clustering objective function as the fitness function, the initial cluster center update process is the continuous iteration process of the spotted kingfisher population, taking the minimum clustering objective function value as the optimal fitness function value; using chaotic mapping to initialize the spotted kingfisher population, setting the current number of iterations to t, and the position of the e-th spotted kingfisher individual in the t-th iteration spotted kingfisher population to C e (t), the position of the g-th individual of the Pied Kingfisher population in the t-th iteration is C g (t), d 1 represents a random number between the interval [0, 1], the chaotic parameter is φ, the number of chaotic sequence particles is c, when C e When (t)<0.5, the position of the e-th Pied Kingfisher in the t+1th iteration When C e When (t)≥0.5, the position of the e-th Pied Kingfisher in the t+1th iteration

[0085] The Pied Kingfisher population enters the exploration phase, which includes the perching strategy and the hovering strategy. The normally distributed random number between the interval [1, j] is set as d 2 , the maximum number of iterations is T,d 3 Represents a random number between the interval [0, 1], habitat parameter The position C of the e-th individual of the Pied Kingfisher in the t+1th iteration of the roosting strategy e (t+1)=C e (t)+(2d 2 -1)·γ 1 ·(C g (t)-C e (t)); calculate the fitness function values ​​of the e-th and g-th individual of the spotted kingfisher population in the t-th iteration, denoted as f e and f g , hover parameter The position C of the e-th individual of the Pied Kingfisher in the t+1th iteration when using the circling strategy e (t+1)=C e (t)+(2d 2 -1)·γ 2 ·(C g (t)-C e (t)); Use dynamic probability switching strategy to balance the proportion of perching strategy and hovering strategy, dynamic probability By comparing the dynamic probability and the random number d 3 Dynamically switch between perching strategy and hovering strategy;

[0086] S312, at this time, calculate the fitness function value corresponding to the individual spotted kingfisher in the spotted kingfisher population, find the current best fitness function value as the current optimal clustering center, and continue to optimize the current optimal clustering center; the spotted kingfisher population enters the development stage, the spotted kingfisher population begins to prey, introduces the triangle wandering strategy to update the position of the spotted kingfisher individual, and calculates the distance l between the e-th spotted kingfisher and the g-th spotted kingfisher individual in the t-th iteration of the spotted kingfisher population 1 , set d 4 Represents a random number between the interval [0, 1], with a walking step length of l 2 = l 1 ·d 4 , walk parameter η = l 1 2 +l 2 2 -2l 1 ·l 2 ·cos(2πd 4 ), for position C e (t+1) is updated, at this time C e (t+1)=C e (t)+η·d 4 ; The Pied Kingfisher population enters the escape stage. Calculate the unpredated efficiency of the Pied Kingfisher population at this time, and set d 5 Represents a random number between the interval [0, 1]. When the random number d 5 When it is greater than the unpredated efficiency, the position C is updated again according to the hunting ability of the individual kingfisher e (t+1), otherwise the position is not updated, and the final position of the individual of the spotted kingfisher in the t+1th iteration is obtained, and the next generation of the spotted kingfisher population is generated and iterated until the current number of iterations reaches the maximum number of iterations, then the iteration is stopped, and the final position of the spotted kingfisher is obtained. The individual position of the spotted kingfisher corresponding to the best fitness function value is recorded as the optimized cluster center;

[0087] S32, replacing the initial cluster center with the optimized cluster center, clustering the inventory time points according to the spatial distance, continuously updating the initial cluster center until the position of the cluster center no longer changes, obtaining the final cluster center, forming a cluster with the final cluster center, and obtaining a new cluster;

[0088] S4. Perform a final check on the inaccurate data based on the new cluster, output the inaccurate inventory data, divide the suspected inventory data set, and complete the accuracy check;

[0089] The S4 comprises the following steps:

[0090] S41. Set the number of the new clustering clusters as h. The new clustering clusters contain several inventory time points. Set a distance threshold, calculate the distances between the final clustering centers and the inventory time points in the new clustering clusters. When the distance between the final clustering center and an inventory time point is greater than the distance threshold, mark the corresponding inventory time point as an inaccurate inventory time point, and identify the inventory suspected data corresponding to the inaccurate inventory time point in the inventory suspected data set to obtain inaccurate inventory data. Divide the inventory suspected data set into an inaccurate inventory data set and an accurate inventory data set to complete the accuracy verification of the inventory data.

[0091] Embodiment 2

[0092] This embodiment also discloses a system for the accuracy verification method of inventory data applied to the secondary development of financial cloud, specifically including: a data missing filling module, a preliminary inaccurate data verification module, a clustering center optimization module, and a final inaccurate data verification module;

[0093] The data missing filling module is used to fill the missing data in the inventory data set based on data correlation;

[0094] The preliminary inaccurate data verification module is used to preliminarily verify the inaccurate data in the inventory suspected data by using an improved clustering algorithm;

[0095] The clustering center optimization module is used to optimize the initial clustering centers by using a multi-strategy fusion pied kingfisher optimization algorithm;

[0096] The final inaccurate data verification module is used to identify the inaccurate inventory data in the inventory suspected data set to complete the accuracy verification.

[0097] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the 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 a suitable manner in any one or more embodiments or examples.

[0098] The preferred embodiments of the invention disclosed above are only used to help illustrate the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. This specification selects and specifically describes these embodiments to better explain the principle and practical application of the invention, so that those skilled in the art in the relevant technical field can well understand and utilize the invention.

Claims

1. The inventory data accuracy verification method applied to the secondary development of financial cloud is characterized by: The steps include: S1. Obtain inventory-related data and outbound delivery-related data in the financial cloud to obtain an initial inventory data set and an initial outbound delivery data set, and then fill in missing data based on data correlation to obtain a processed inventory data set and a processed outbound delivery data set; S2. Obtaining a suspected inventory data set based on the processed inventory data set and the processed outbound data set, and performing a preliminary check on inaccurate inventory data in the suspected inventory data set using an improved clustering algorithm in combination with spatial distance and time series to obtain a clustering objective function; S3, taking the clustering objective function as a fitness function, using an optimization algorithm to optimize the initial clustering center of the improved clustering detection algorithm, obtaining an optimized clustering center, and generating a new cluster; S4. Perform a final check on the inaccurate data based on the new cluster, output the inaccurate inventory data, divide the suspected inventory data set, and complete the accuracy check.

2. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 1 is characterized in that: The S1 comprises the following steps: S11. Obtaining inventory-related data and delivery-related data in the financial cloud to form an initial inventory data set and an initial delivery data set; S12: Fill in missing data based on the data correlation between the initial inventory data set and the initial outbound data set to obtain a processed inventory data set and a processed outbound data set.

3. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 2 is characterized in that: The S12 comprises the following steps: S121, calculating the data correlation of the inventory data according to the inventory time series in the initial inventory data set; and then calculating the data correlation of the outbound data according to the outbound time series in the initial outbound data set; S122, calculating the Euclidean distance of the inventory data, and combining the data correlation of the inventory data to obtain the inventory correlation distance; then calculating the Euclidean distance of the outbound data, and combining the data correlation of the outbound data to obtain the outbound correlation distance; For the missing inventory data in the initial inventory data set, data missing is filled in combination with inventory related distances and neighboring inventory data of the missing inventory data to obtain a processed inventory data set; For the missing outbound data in the initial outbound data set, the outbound related distance and the outbound data adjacent to the missing outbound data are combined to obtain a processed outbound data set.

4. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 3 is characterized in that: The S2 comprises the following steps: S21, obtaining a warehouse-in and warehouse-out data set, accumulating corresponding data in the processed warehouse data set and the processed warehouse-out data set to obtain a data set to be verified; Set the verification threshold, compare the data set to be verified with the inbound and outbound data set, and obtain the suspected inventory data set; S22. Combining spatial distance and time series, using an improved clustering algorithm to perform preliminary verification on the inaccurate inventory data in the suspected inventory data set, and obtaining a clustering objective function.

5. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 4 is characterized in that: The S22 comprises the following steps: S221, selecting any inventory time point in the inventory suspected data set as an initial cluster center, constructing a topological space, and calculating the spatial distance between the initial cluster center and other inventory time points in the topological space to obtain a new cluster center; S222: Calculate the spatial distance between the new cluster center and the most recent inventory time point, generate an updated cluster center, identify inaccurate inventory data in the inventory suspected data set, complete a preliminary check on the inaccurate data, and establish a clustering objective function.

6. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 5 is characterized in that: The S221 includes the following steps: S2211, calculating the spatial distance from the initial cluster center to other inventory time points to obtain the closest spatial distance, and recording the inventory time point corresponding to the closest spatial distance as the closest spatial distance inventory time point; S2212: Record the closest spatial distance inventory time point as a new cluster center, and use the new cluster center to replace the initial cluster center.

7. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 6 is characterized in that: The S3 comprises the following steps: S31, taking the clustering objective function as the fitness function, introducing chaos mapping, triangle walk strategy and dynamic probability switching strategy to improve the piebald kingfisher optimization algorithm, obtaining a multi-strategy fusion piebald kingfisher optimization algorithm, and using the multi-strategy fusion piebald kingfisher optimization algorithm to optimize the initial cluster center of the improved cluster detection algorithm, to obtain an optimized cluster center; S32, replacing the initial cluster center with the optimized cluster center, clustering the inventory time points according to the spatial distance, continuously updating the initial cluster center until the position of the cluster center no longer changes, obtaining the final cluster center, forming a cluster with the final cluster center, and obtaining a new cluster.

8. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 7 is characterized in that: The S31 comprises the following steps: S311, setting a search space with dimension j, in which there is a population of spotted kingfishers, taking the clustering objective function as the fitness function, the initial cluster center update process is the continuous iteration process of the spotted kingfisher population, taking the minimum clustering objective function value as the optimal fitness function value; using chaotic mapping to initialize the spotted kingfisher population, setting the current number of iterations to t, and the position of the e-th spotted kingfisher individual in the t-th iteration spotted kingfisher population to C e (t), the position of the g-th individual of the Pied Kingfisher population in the t-th iteration is C g (t), d1 represents a random number between the interval [0, 1], the chaotic parameter is φ, the number of chaotic sequence particles is c, when C e When (t)<0.5, the position of the e-th Pied Kingfisher in the t+1th iteration When C e When (t)≥0.5, the position of the e-th Pied Kingfisher in the t+1th iteration The Pied Kingfisher population enters the exploration phase, which includes the roosting strategy and the hovering strategy. The normally distributed random number between the interval [1, j] is set as d2, the maximum number of iterations is T, d3 represents a random number between the interval [0, 1], and the roosting parameter The position C of the e-th individual of the Pied Kingfisher in the t+1th iteration of the roosting strategy e (t+1)=C e (t)+(2d2-1)·γ1·(C g (t)-C e (t)); calculate the fitness function values ​​of the e-th and g-th individual of the spotted kingfisher population in the t-th iteration, denoted as f e and f g , hover parameter The position C of the e-th individual of the Pied Kingfisher in the t+1th iteration when using the circling strategy e (t+1)=C e (t)+(2d2-1)·γ2·(C g (t)-C e (t)); Use dynamic probability switching strategy to balance the proportion of perching strategy and hovering strategy, dynamic probability Dynamically switch between perching strategy and hovering strategy by comparing dynamic probability and random number d3; S312. At this time, the fitness function value corresponding to the individual spotted kingfisher in the spotted kingfisher population is calculated, the current best fitness function value is found as the current optimal clustering center, and the current optimal clustering center is continuously optimized; the spotted kingfisher population enters the development stage, the spotted kingfisher population begins to prey, and the triangle wandering strategy is introduced to update the position of the spotted kingfisher individuals, and the distance l1 between the e-th spotted kingfisher and the g-th spotted kingfisher individuals in the t-th iteration of the spotted kingfisher population is calculated, d4 ​​is set to represent a random number between the interval [0, 1], the wandering step length l2 = l1·d4, and the wandering parameter η = l1 2 +l2 2 -2l1·l2·cos(2πd4), for position C e (t+1) is updated, at this time C e (t+1)=C e (t)+η·d4; the Pied Kingfisher population enters the escape stage, and the unpreyed efficiency of the Pied Kingfisher population is calculated at this time. Set d5 to represent a random number between the interval [0, 1]. When the random number d5 is greater than the unpreyed efficiency, the position C is updated again according to the hunting ability of the Pied Kingfisher individual. e (t+1), otherwise the position is not updated, and the final position of the individual of the spotted kingfisher in the t+1th iteration is obtained, and the next generation of the spotted kingfisher population is generated to continue the iteration until the current number of iterations reaches the maximum number of iterations, then the iteration is stopped to obtain the final position of the spotted kingfisher, and the individual position of the spotted kingfisher corresponding to the best fitness function value is recorded as the optimized cluster center.

9. The inventory data accuracy verification method applied to the secondary development of financial cloud according to claim 8 is characterized in that: The S4 comprises the following steps: S41. Set a distance threshold, calculate the distance between the final cluster center and the inventory time point in the new cluster, compare it with the distance threshold, output inaccurate inventory data, divide the suspected inventory data set into an inaccurate inventory data set and an accurate inventory data set, and complete the accuracy check.

10. A system for implementing the inventory data accuracy verification method applied to financial cloud secondary development as claimed in any one of claims 1 to 9, characterized in that: Specifically include: Data missing filling module, inaccurate data preliminary verification module, cluster center optimization module and inaccurate data final verification module; The data missing filling module is used to fill the data missing of the inventory data set based on data relevance; The inaccurate data preliminary verification module is used to perform preliminary verification on inaccurate data in the suspected inventory data using an improved clustering algorithm; The cluster center optimization module is used to optimize the initial cluster center using a multi-strategy fusion Kingfisher optimization algorithm; The inaccurate data final verification module is used to identify inaccurate inventory data in the inventory suspected data set and complete accuracy verification.

Images