197 results about "Dynamic storage" patented technology

The invention discloses a cargo warehouse-in and warehouse-out management method and management system, and relates to the technical field of information management. Through dynamic storage position optimization, the warehouse-in and warehouse-out frequency, the associated sales relationship and the storage demand characteristics of cargos are combined with a warehouse three-dimensional space model; and generating a cargo storage location allocation scheme by using a multi-objective optimization algorithm. According to the scheme, the picking time of high-frequency warehouse-in and warehouse-out commodities is remarkably shortened, the average time consumption of single-time commodity picking is reduced, meanwhile, the combined picking error rate of associated commodities is reduced, the problem that in the prior art, storage positions are disjointed with warehouse-in and warehouse-out requirements is solved, through predictive process scheduling, cross congestion of operation channels is effectively avoided, and the working efficiency is improved. Therefore, smooth operation of the in-out warehouse process is realized, the occurrence frequency of congestion events of the in-out warehouse channel is reduced, the delay time is shortened, and the problem of congestion of the in-out warehouse process in the prior art is solved.

The invention relates to the technical field of smart community data management, in particular to a smart community metadata interaction method and system based on an edge computing framework, and the method comprises the steps: carrying out the dynamic storage of multi-dimensional data through a hierarchical rolling cache; generating a uniform resource vector through sequential progressive mapping; community semantic tags are added for the uniform resource vectors through extensible tag slots, connection is established according to association rules, and a resource semantic graph oriented to community services is constructed; receiving a resident task request, constructing a multi-factor weighted scoring engine based on the resource semantic graph, and generating a task data packet with a priority label; a two-channel hybrid scheduling network is constructed, a first channel outputs resource availability and node association degree, and a second channel outputs time sequence characteristics and load prediction parameters; and fusing dual-channel output to obtain task-node correlation and generate an optimal matching strategy. According to the method, task collaboration and resource adaptive management are realized through semantic resource modeling and hybrid scheduling.

The invention relates to the technical field of enterprise office collaboration, and discloses an enterprise series intelligent office collaboration data processing system based on edge computing. The system comprises a collaborative data acquisition module, a collaborative feature analysis module, an edge task scheduling module and a dynamic storage optimization module. The collaborative data acquisition module is deployed at a local edge node of an enterprise, captures a collaborative operation behavior flow of the multi-source office terminal in real time, and extracts space-time distribution features to generate an original collaborative feature vector; the collaborative feature analysis module analyzes the coupling strength of related data through a multi-dimensional correlation analysis engine, quantifies the frequency fluctuation amplitude of a communication session, and fuses to generate a collaborative efficiency quantized value; the edge task scheduling module dynamically divides task load levels based on the value, and matches execution priorities to generate a distributed calculation task allocation strategy; and the dynamic storage optimization module responds to the strategy, monitors the throughput delay of the storage nodes, reconstructs a high-frequency access data storage index, and generates a cross-node storage topology configuration.

The invention relates to the technical field of data storage, and discloses an intelligent hierarchical storage method and system for cold and hot data of a solid state disk, and the method comprises the steps: collecting spatial position change data and vibration state data, and carrying out the preprocessing to obtain a multi-dimensional feature data set; performing quantification processing on the influence degree of the spatial position change data and the vibration state data to obtain an interference weight value; constructing a joint feature matrix based on the interference weight value and the historical access record, and predicting hotspot data distribution; if the matching degree of the hotspot data distribution and the current storage layout is lower than a preset matching degree threshold value, migrating the high-frequency access data to a low-delay area to obtain an optimized storage layout; when the access response time is larger than a preset time threshold value, log data are generated, and delay load features are extracted; and carrying out secondary adjustment on the optimized storage layout to obtain a final dynamic storage scheme. According to the method, dynamic optimization of the storage performance of the solid state disk can be realized, and delay loss caused by cold and hot data mismatching is reduced.

The invention provides a time sequence data storage method and system, electronic equipment and a storage medium, and the method comprises the steps: obtaining time sequence data and metadata information of the time sequence data, the metadata information comprising a first point location identifier, a data source and a data type corresponding to the time sequence data; according to the data source and the data type, determining a storage strategy of the time series data, including a target storage period and a target storage position; according to the storage strategy and the first point position identification, a target storage instance corresponding to the time sequence data is matched from a storage instance set, all storage instances in the storage instance set are combined and constructed according to different storage strategies and the point position identification, and all the storage instances are parallel storage instances; and storing the time sequence data to a time sequence database through the target storage instance. By determining a dynamic storage strategy driven by metadata and a parallel storage instance, the problem of performance bottleneck during high-concurrency writing of massive time series data is solved, and the real-time performance and reliability of time series data storage are ensured.

The invention relates to the technical field of cloud storage, can be applied to business system platforms of financial science and technology, medical treatment and health and the like, and discloses a hierarchical storage method, device and equipment for data files and a medium, and the method comprises the following steps: acquiring the data files in a storage unit, and extracting data characteristic parameters in the data files; performing multi-dimensional quantitative analysis on the data feature parameters to obtain a dimensional feature analysis value; identifying a dynamic weight of the data feature parameter, and performing weighted calculation on the data feature parameter according to the dimension feature analysis value and the dynamic weight to obtain a dynamic storage analysis value; identifying a data storage type label according to the dynamic storage analysis value; matching the data storage type label with strategies in a storage strategy library to obtain a target storage strategy; determining a storage hierarchy according to the target storage strategy; a storage address of the data file in the storage hierarchy is identified and the data file is stored in the storage address. The accuracy of hierarchical storage of the data files can be improved.

The invention discloses a data real-time analysis and fault early warning method and system based on a digital electric port box transformer substation, and belongs to the technical field of power equipment monitoring and control systems. The method comprises the following steps: acquiring environmental parameters of each component of the box transformer substation in real time through a monitoring network, and performing data smooth denoising to obtain a parameter sequence; constructing a dynamic storage framework based on the sequence in combination with space-time attributes; extracting historical records of the target component to perform multi-dimensional sorting and change rate calculation, and identifying potential anomalies; when the change rate exceeds the limit and the distribution boundary breaks through, generating an alarm sequence and determining a preliminary abnormal mark; comparing the mark with the historical record, fusing the influence weight analysis verification accuracy, and obtaining a final early warning grading result; and parameter summary information is pushed to a control system according to a grading result, response feedback is obtained, and closed-loop control of the operation state is formed. According to the invention, the problem of control lag caused by data noise and delay is solved, and the timeliness, reliability and automation level of box transformer substation control are improved.

The invention belongs to the field of computer system structures and artificial intelligence accelerators, and discloses a software and hardware collaborative heterogeneous storage computing accelerator oriented to full-chip DLRM reasoning. A control domain composed of a static random access memory and an embedded dynamic random access memory, an on-chip storage domain, an in-memory computing domain and an on-chip computing domain are integrated in the same chip, an embedded table is compressed by introducing a hierarchical clustering quantization frame, and a dynamic storage management strategy of access frequency perception is combined. The embedded access path is shortened, and the bandwidth requirement is reduced. According to the method, the reasoning process of the whole set of DLRM can be independently completed, and the memory access overhead in a traditional terminal SoC is remarkably reduced by reducing cross-domain data migration between the calculation unit and the storage unit, so that the model reasoning performance and energy efficiency are improved.

The invention relates to the technical field of data processing, in particular to an intelligent data storage and knowledge base integrated management system enabling an AI large model. The system comprises a data hierarchical storage module, a hybrid index construction module, an integrated knowledge base establishment module and an integrated knowledge base index module. Firstly, a data value score is calculated based on a multi-dimensional evaluation index, and a dynamic storage grading mechanism is established; secondly, a three-layer mixed index structure is adopted to index different types of hierarchical data, storage positions are adjusted, and a multi-source heterogeneous data storage layer is obtained; fusing the entity relationship to generate a triple, constructing a dynamic knowledge graph, and introducing an incremental updating mechanism to update in real time to obtain an integrated knowledge base; and finally, inputting a natural language, performing semantic analysis, obtaining a multi-modal index result by adopting a mixed index, and establishing an integrated knowledge base self-diagnosis mechanism. According to the method, the multi-source heterogeneous data is analyzed and processed, so that the purposes of data storage and knowledge base integrated management are achieved.

The invention provides a memory management method and a memory controller. The memory management method is applied to the storage device provided with a memory module and a buffer memory. The method comprises the steps of determining a necessary function and a high-frequency call function as a first type function by analyzing a call relationship of functions in a firmware program, and pre-loading the first type function into a resident storage area of a buffer memory; dividing the rest of the second type functions into a plurality of storage libraries according to the function calling relationship, and storing the rest of the second type functions in a storage module; and when the specific function needs to be executed, dynamically loading the storage library containing the target function into the dynamic storage area of the buffer memory. According to the method, through weight calculation and an intelligent allocation strategy, the number of times of switching of a storage library (Bank) can be reduced, so that the resource utilization rate of a memory is improved, the overall performance of a system is improved, automatic management of firmware codes is achieved, and the complexity and error risk of manual adjustment are reduced.

The invention discloses an intelligent storage automatic control optimization method and system based on dynamic environment perception, and relates to the technical field of intelligent storage control. According to the method, a specific scene type of a specified warehouse is firstly obtained, a corresponding dynamic perception allocation scheme is determined through scene matching, and then a collaborative operation association object is determined, so that accurate adaptation of dynamic environment perception and a warehouse scene is realized; then, collecting dynamic environment sensing data of a storage area where a collaborative operation association object is located, and after collaborative association processing is carried out, generating a structured fusion data set adapted to different time sequence scenes; dynamic storage parameter calibration is carried out, an initial control strategy is generated, and effectiveness evaluation is completed; whether the path planning scheme and the operation execution sequence are dynamically adjusted or not is decided according to the evaluation result, dynamic adaptation and accurate regulation and control of intelligent storage automatic control are achieved, and the adaptability of storage control, the smoothness of process operation and the stability of operation execution in a complex dynamic scene are effectively improved.

The invention relates to the technical field of electronic design automation, in particular to a function coverage rate data compression and storage method, system and device and a storage medium. Collecting the coverage rate of each coverage point or cross coverage and the hit times of each bin subordinate to the coverage point or the cross coverage; acquiring a coverage rate grade and a corresponding data structure; dynamically selecting a target data structure according to the target coverage rate grade of the coverage rate of the target coverage point or the cross coverage; operating a corresponding target data structure through a dynamic storage selector according to the target data structure, wherein the target data structure comprises a flag bit of the current data structure and storage structure data; wherein the storage structure data comprises at least one of the serial number of the bin and the hit frequency of the bin, and the storage overhead is remarkably reduced.

The invention relates to the technical field of intelligent warehouse management, in particular to a warehouse operation method and system based on storage location and picking path integrated optimization, and the method comprises the steps: building a dynamic storage location distribution optimization model with the minimization of the total space waste value of a whole warehouse as an optimization target based on an initialized storage layout parameter set; performing iterative solution on the dynamic storage location allocation optimization model to obtain an initial tray storage location allocation scheme; according to the real-time order information and the initial tray storage location distribution scheme, generating an optimal order picking path of each order; and carrying out storage location secondary allocation on the remaining material storage location according to the warehouse remaining material storage state after the optimal order picking path is executed to obtain a storage location reallocation scheme, thereby generating a storage location-path integrated operation instruction, and controlling the warehouse management system to operate through the storage location-path integrated operation instruction. According to the invention, through collaborative optimization of the storage location allocation and the picking path, integrated collaborative optimization of the storage layout, the storage location allocation and the picking path is realized.

The invention discloses a time sequence data partition storage method for a high-frequency acquisition system, relates to the technical field of data storage management, and is used for solving the problem of dynamic unbalance of a high-frequency write-in structure. According to the method, a time sequence data dynamic storage mechanism based on write-in pressure sensing and partition structure linkage is constructed, rapid response and structure self-adaptive adjustment of burst write-in loads in a high-frequency acquisition system are achieved, the write-in overload state is recognized by constructing a sliding write-in rate sequence and a write-in pressure factor, and therefore the write-in pressure sensing and partition structure self-adaptive adjustment of the high-frequency acquisition system is achieved. The method comprises the following steps: dynamically triggering a structure splitting operation of a main partition, generating a sub-partition with a time boundary and an independent buffer path, judging and realizing a load balancing route of a write-in request by adopting timestamp mapping and a buffer area utilization rate, generating a structure merging evaluation factor in combination with write-in activeness and delay statistics, and dynamically adjusting a partition structure according to trend change, and the write-in structure and the system load are cooperatively matched, so that the structural elasticity and the write-in efficiency of the system in a high-concurrency scene are improved.

The invention relates to the technical field of data processing, and discloses a compiling method and system for generating hardware accelerator executable codes facing a multi-mode large language model, a terminal and a storage medium. The method comprises the following steps: constructing a computational graph according to the structure of a to-be-deployed large language model and operators supported by a hardware platform; traversing the computational graph to deduce output tensor type information of each operator, and verifying the type consistency between adjacent operators; based on the verified type information, executing static memory allocation to determine a static storage address of the weight data, and executing dynamic memory allocation to determine a dynamic storage address of the state data; then combining the verified type information and the static and dynamic storage addresses to generate an intermediate representation instruction sequence for configuring a hardware acceleration platform register; and finally, converting the intermediate representation instruction sequence into a target code file which can be directly loaded and executed by a hardware acceleration platform. The compiling efficiency, the deployment reliability and the resource utilization rate are improved.

The invention relates to the technical field of warehouse management, in particular to a shoe and clothing warehouse management method and system based on dynamic inventory and intelligent boxing. According to the invention, the basic attribute data and the historical operation data of the stored shoe and clothing commodities are analyzed to calculate the commodity fluidity comprehensive score, and the commodity fluidity grade is divided according to the score, so that the sales speed and the turnover characteristic of the commodities can be fully considered in the storage location distribution; the dynamic and adaptive adjustment of storage location allocation is realized by collecting the storage actual picking efficiency and storage location utilization rate data in real time and combining with a dynamic adjustment model to carry out iterative optimization on an initial storage location allocation scheme, and the dynamic storage location allocation is combined with an intelligent boxing scheme. The boxing scheme is generated according to the commodity size and the storage location space parameters, the warehousing operation instruction is automatically generated, automatic coordination of links such as commodity storage, picking and boxing is achieved, the warehousing operation efficiency is improved, misoperation and logistics cost are reduced, meanwhile, the cargo carrying path is optimized, and warehousing resource waste is reduced.

The invention provides a data processing method, electronic equipment and a vehicle, and relates to the technical field of data processing, and the method comprises the steps: obtaining the correlation evaluation data of each stored memory data, carrying out the multi-dimensional value evaluation of the memory data according to the correlation evaluation data, and determining a value evaluation score. And determining a corresponding forgetting strategy according to the value evaluation score, and performing storage management on the memory data according to the forgetting strategy. Different forgetting strategies are adopted for storage management of data with different evaluation values, high-value data are reserved, low-value data are deleted or compressed, part of storage space is released, and continuous expansion of the storage space is avoided. In the subsequent data retrieval process, the retrieval efficiency of the memory data can also be improved, and continuous reduction of the system efficiency is avoided. Through dynamic storage management of the memory data in the method, self-optimization and evolution of the memory management system are realized, and the requirements of users for data management are met.

This invention provides a data storage method and system for tuberculosis detection, relating to the field of tuberculosis detection technology. The data storage method for tuberculosis detection includes the following steps: S1. Collecting whole-genome sequencing data of Mycobacterium tuberculosis, host serum IgG titer, and drug sensitivity test results; S2. Calculating genetic distance D based on a reverse evolution model to generate four-dimensional spatiotemporal coordinates (t, x, y); S3. Performing data partitioning and storage based on the drug target barrier value β; S4. Generating dynamic metadata using a host-pathogen dynamics model and compressing and storing it. This invention implements a dynamic storage entropy adjustment algorithm at the hardware and software collaborative level, continuously optimizing the matching efficiency of data compression and physical storage. This results in an intelligent data hub that can perceive the evolutionary pulse of pathogens and autonomously optimize resources, providing support for clinical tuberculosis prevention and control decisions with temporal depth, spatial correlation, and risk evolution.

The invention provides a memory information processing method of an electric energy meter, electronic equipment and a storage medium. The method comprises the following steps: acquiring original working data of the electric energy meter in real time and determining a data type of the original working data; converting the original working data into to-be-stored data of a target type; according to the target type and the data type, a target storage position is allocated for the to-be-stored data in a memory of the electric energy meter, the to-be-stored data is stored in the target storage position, the memory of the electric energy meter comprises a static storage area and a dynamic storage area, and the static storage area is used for storing the to-be-stored data in the target storage position. The dynamic storage area comprises a first storage sub-area and a second storage sub-area, and the storage capacity of each storage unit in the first storage sub-area is smaller than that of each storage unit in the second storage sub-area. The memory utilization rate is improved, and more memory fragments are effectively prevented from being generated.

The invention discloses a model increment training method and device based on dynamic storage, a storage medium and a program product, and relates to the technical field of artificial intelligence, and the model increment training method based on dynamic storage comprises the steps: obtaining at least one piece of original sample data, and determining an information amount score corresponding to the original sample data; determining the original sample data with the information content scores higher than a preset score threshold as training sample data, storing the training sample data in a preset memory, and accumulating the information content scores of the training sample data in the preset memory to obtain a corresponding accumulated information content score; and performing incremental training on the to-be-trained model by utilizing the training sample data in the preset memory to obtain a target model in response to the condition that the accumulated information amount score or the sample number corresponding to the training sample data in the preset memory meets a preset triggering condition. The technical problem that the training effect of a traditional field small model is affected due to data distribution offset of training data is solved.

The embodiment of the invention discloses a video data management method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a video code stream from each video channel, and compressing the obtained video code stream; according to the historical access data and the scene label corresponding to each video channel, configuring a dynamic storage period corresponding to each video channel; according to the dynamic storage period corresponding to each video channel, storing the compressed video code stream into a target storage area; detecting the residual storage space of the target storage area; when the residual storage space is smaller than a preset early warning storage threshold value, the target video code stream is deleted according to the access frequency of each video code stream and the dynamic storage period corresponding to each video channel, and according to the scheme, storage resource distribution can be optimized, storage occupation can be reduced, and key videos are prevented from being deleted by mistake.

The application relates to an inorganic chloride-doped polymer electrolyte hybrid dielectric film material and a preparation method and application thereof, and belongs to the technical field of dielectric film materials and devices. The inorganic chloride-doped polymer electrolyte hybrid dielectric film material has significantly enhanced capacitance and ion migration capacity, can be used for developing an organic transistor memory with multi-level response to different pulse voltages, and can realize a wide memory window of 0.548 volts at a low working voltage of-1.5 volts. The memory also has good static and dynamic storage characteristics, and can realize accurate regulation and identification of different sizes of memory windows through programming gate pulse voltage. The novel polymer electrolyte hybrid dielectric film material and the organic transistor memory with multi-level storage characteristics thereof have the advantages of simple structure, excellent performance, easy scale-up manufacturing and the like, and will have good popularization and application prospects in the field of low-power-consumption optoelectronic devices and storage technologies.

The application discloses a high-density stereoscopic warehouse dynamic storage space allocation method for WMS, and relates to the technical field of artificial intelligence. Stereoscopic warehouses are arranged along the direction of the roadway, and stackers, shuttles and hoists are arranged in the stereoscopic warehouses. Field programmable gate array acceleration cores are arranged in the edge nodes of the stereoscopic warehouses. The acceleration cores are interconnected with the WMS through a time-sensitive network. The method comprises the following steps: step 1, the acceleration core constructs a beat dependency graph based on the current carrying instruction issued by the WMS, wherein the carrying motion is taken as an event node, and the action connection relationship is taken as a directed edge; step 2, the acceleration core constructs a colored Petri net; and step 3, the acceleration core constructs a graph attention reinforcement learning decision module and takes a heterogeneous graph as input, and the WMS drives the corresponding carrying equipment to perform the actions of warehousing, moving and replenishment. The application can significantly improve the system throughput, reduce the equipment idle rate and beat jitter under the sub-millisecond end-to-end decision delay.

One or more techniques and / or systems are provided for dynamically provisioning logical storage pools of storage devices for applications. For example, a logical storage pool, of one or more storage devices, may be constructed based upon a service level agreement for an application (e.g., an acceptable latency, an expected throughput, etc.). Real-time performance statistics of the logical storage pool may be collected and evaluated against the service level agreement to determine whether a storage device does not satisfy the service level agreement. For example, a latency of a storage device within the logical storage pool may increase overtime as log files and / or other data of the application increase. Accordingly, a new logical storage pool may be automatically and dynamically defined and provisioned for the application to replace the logical storage pool. The new logical storage pool may comprise storage devices expected to satisfy the storage level agreement.

The application provides a memory information processing method of an electric energy meter, an electronic device and a storage medium. The method comprises the following steps: acquiring original working data of the electric energy meter in real time and determining a data type of the original working data; converting the original working data into to-be-stored data of a target type; and according to the target type and the data type, allocating a target storage position for the to-be-stored data in a memory of the electric energy meter and storing the to-be-stored data into the target storage position. The memory of the electric energy meter comprises a static storage area and a dynamic storage area, the dynamic storage area comprises a first storage sub-area and a second storage sub-area, and the storage capacity of each storage unit in the first storage sub-area is smaller than the storage capacity of each storage unit in the second storage sub-area. The memory utilization rate is improved, and the generation of more memory fragments is effectively avoided.

This invention discloses a dynamic data area sharing system and method for Flash chips, belonging to the field of computer firmware technology. The system includes a BIOS core framework, a shared storage component, and a dynamic shared storage area. The BIOS loads and runs the shared storage component through the BIOS core framework. The shared storage component can dynamically use the shared storage area according to the needs of the current computer boot, performing BIOS upgrade backups and BIOS log recording. This invention solves the problem of BIOS backup and BIOS log storage by dynamically dividing the SPI Flash memory into a storage area when the BIOS space on the computer motherboard is limited.

The invention relates to the technical field of medicine storage, and discloses an AI-based coenzyme Q10 soft capsule dynamic storage environment control system, which comprises a data acquisition module used for acquiring temperature data, humidity data, illumination intensity data and medicine state data; the data preprocessing module is used for cleaning the collected data, removing abnormal values and noise data, and performing normalization processing to obtain preprocessed data; the model prediction module is used for inputting the preprocessed data into an LSTM (Long Short Term Memory) model, predicting the change trend of environmental parameters in a future time window, simulating a degradation path of the activity of the coenzyme Q10 and outputting a prediction result; the environment regulation and control module is used for generating a control strategy according to the prediction result and realizing dynamic adjustment of environment parameters through an execution mechanism; according to the invention, the dynamic and accurate regulation and control of the storage environment are realized, the manual intervention is reduced, the labor cost is reduced, and the overall efficiency of storage management is improved.

The invention discloses an in-vehicle active noise control method and system based on a convolutional recurrent network, and the method comprises the steps: obtaining reference signals collected by a sensor in real time, forming a cache data set by M reference signals, and enabling reference data in the cache data set to slide rightwards for a jump length after a short-time Fourier transform sliding window; the reference signals in the cache data set are subjected to short-time Fourier transform according to algorithm updating frequency, and one frame of time-frequency data is formed in each time of short-time Fourier transform; storing the N frames of time-frequency data by adopting a first-in first-out principle; and inputting the N frames of time-frequency data into the trained convolutional cyclic network model, and outputting an output signal for controlling a loudspeaker according to inverse Fourier transform. According to the method, the recurrent neural network is used for dynamically storing and updating time sequence steps, the noise reduction capability of the depth model is reserved, and the requirement for real-time noise reduction of the whole vehicle is met.

The invention discloses an intelligent warehouse dynamic storage location optimization system and method based on AI and digital twinning, and relates to the technical field of storage location optimization. Obtaining current inventory state information of the goods shelf; establishing a nearest goods storage distance constraint function, a goods shelf bearing capacity constraint function and a goods shelf stability constraint function based on the distance between each storage location and the warehouse entrance and the weight of the unit of goods stored in each storage location, and establishing a target function; initializing a particle swarm based on the number of units of goods to be stored and the current inventory state information of each goods shelf; solving the particle population by using a genetic algorithm to obtain an optimal particle individual; based on the optimal particle individuals, determining a storage location allocation scheme of each to-be-stored cargo unit; and establishing a digital twinborn model of the warehouse, and highlighting the storage location model of each to-be-stored cargo unit in the digital twinborn model based on the storage location allocation scheme of each to-be-stored cargo unit. The warehouse-in and warehouse-out efficiency can be ensured, and meanwhile, the goods storage safety is considered.

The application relates to the technical field of big data storage and management, and provides a dynamic storage management method and a management platform for Internet of Things equipment data. The method comprises the following steps: in response to a data query request, calculating a target storage optimization index for target combined data to obtain a target storage optimization index; determining a first database for storing the target combined data based on the target storage optimization index; disassembling the target combined data to obtain a plurality of target single combined data, calculating a storage optimization index of each target single combined data to determine a storage database of each target single combined data; migrating the target single combined data with a storage optimization level higher than that of the first database to the corresponding storage database, and deleting the target single combined data with a storage optimization level not higher than that of the first database; and migrating the target combined data to the first database to realize dynamic storage optimization of the target data combination.