889 results about "Calculation technique" patented technology

The invention relates to the technical field of cloud side end computing, and discloses a cloud side end cooperative computing resource dynamic allocation method based on task-aware migration. The method comprises the following steps: acquiring real-time load characteristics and resource demand characteristics of calculation tasks in a cloud side end system, and dividing task priority queues in combination with task type identifiers; extracting historical execution records of tasks at cloud, edges and terminal nodes, constructing a task execution feature library, and generating a resource demand prediction model in combination with real-time load features; analyzing network transmission time delay characteristics of a cloud end and edge nodes, measuring real-time calculation capability fluctuation data of terminal equipment, and establishing an inter-node resource collaboration degree evaluation matrix; generating an initial migration strategy according to the prediction model and the evaluation matrix, monitoring actual resource occupancy deviation of the task, forming a final decision in combination with a node resource state correction strategy, triggering cross-node migration, and synchronously updating the priority queue and the evaluation matrix.

The invention belongs to the technical field of artificial intelligence and high-performance computing, and particularly relates to a neural network large model efficient reasoning method based on multiple GPGPUs. The method aims to solve the problems of high communication overhead, non-uniform load, low resource utilization rate, high data transmission delay and the like among multiple processors. Dividing a calculation task into a plurality of sub-graphs through static analysis and mixed granularity partitioning of a model calculation graph; distributing the sub-graphs to the optimal GPGPU based on a weighted cost function in combination with heterogeneous resource perception and a dynamic mapping strategy; a global pipeline scheduling plan is constructed by using communication topology perception, and calculation and communication overlap are maximized; data are loaded in advance through a host side hierarchical caching and asynchronous prefetching mechanism, and transmission delay is hidden; multi-stream concurrent execution and event-based lightweight synchronization are adopted on each GPGPU, so that waiting overhead is reduced. According to the method, the reasoning delay can be remarkably reduced, the throughput and the hardware utilization rate are improved, and the method has good adaptivity and expandability.

The invention provides a parallel computing method and device, electronic equipment and a storage medium, and relates to the technical field of parallel computing, and the method comprises the steps: carrying out the first protocol operation of a target tensor based on each computing core in each stream processor cluster, and generating a data block containing the computing result of each computing core; writing a data block generated by each stream processor cluster into a shared cache; under the condition that each stream processor cluster completes the first protocol operation, reading a data block written by each stream processor cluster from the shared cache; and executing a second protocol operation on the data block read from the shared cache to generate a calculation result of the target tensor. According to the method and device provided by the invention, the parallel architecture and memory access characteristics of the artificial intelligence chip can be better matched, the unnecessary calculation delay and synchronization overhead of the cross-flow processor cluster in the parallel calculation process are reduced, the bandwidth utilization rate of the shared cache is improved, and the overall performance and calculation efficiency of parallel calculation are remarkably improved.

The invention relates to a final-stage cache design method based on near memory calculation, and belongs to the technical field of near memory calculation. And compared with a near memory processor near a traditional main memory, data access and calculation can be carried out more quickly, and the calculation throughput rate can be improved. According to the near storage calculation, a near storage processor and a last-stage cache storage array are directly designed in an integrated mode, faster memory access is achieved through an internal bus interconnection mode, and the problem of a storage wall between the calculation speed and the memory access speed and the problem of a power consumption wall of data handling are relieved to a certain degree. The last-stage cache controller cooperates with the embedded processor, and can dynamically coordinate between a standard cache function and a calculation mode. Through a synchronization mechanism in a last-stage cache controller, a memory access request from a host CPU and a calculation task executed by an embedded processor in a last-stage cache are transparently balanced, so that efficient near-memory calculation under cache consistency is realized.

The embodiment of the invention discloses a dynamic computing power scheduling method and device based on deep learning and a medium, belongs to the technical field of cloud computing, and solves the problem that cloud computing resources are seriously wasted due to the fact that a cloud computing scheduling method in the prior art is prone to unbalanced resource allocation. Comprising the following steps: preprocessing acquired cloud computing data to generate a structured feature vector; wherein the cloud computing data at least comprises cloud node state data, computing task attributes and environment data; through a multi-head attention mechanism, task resource space-time correlation features corresponding to the structured feature vectors are extracted; the task resource space-time correlation features are input into a weight dynamic prediction module and a resource matching degree scoring module at the same time so as to carry out parallel processing of weight prediction and matching degree scoring calculation; and solving a multi-objective optimization function corresponding to computing power scheduling based on a parallel processing result, and generating a computing power scheduling instruction based on a solving result.

To accelerate search speeds for approximate nearest neighbor searches of vector databases, compute-in-memory techniques using NAND memory structures are introduced. For each element of the database, a kernel of its M nearest neighbors is determined. For each vector of the database, both the vector and its kernel are programmed in the arrays of a NAND memory based accelerator card, so that the vectors will be written into the memory arrays both as themselves and also in kernels of vectors for which they are a nearest neighbor. Metadata, associating the locations of the kernel members with the correspond vector is also stored in the memory system. After determining the input's nearest neighbor at one level of search, the metadata is then used to locate that nearest neighbor's nearest neighbors and their distances to the input vector are then computed in parallel in a compute-in-memory vector-vector dot product multiplication.

The invention provides an edge calculation optimization method for a seawater parameter monitoring Internet of Things system, and belongs to the technical field of edge calculation, and the method comprises the steps: constructing a seawater parameter monitoring sensor network, and calculating an energy consumption evaluation reference value based on a data change rate and a disorder degree; performing feature extraction on the seawater parameter data; deploying a double-model collaborative architecture comprising a terminal lightweight model and a center complete model; setting a dynamic energy consumption threshold judgment mechanism to allocate calculation tasks; executing lightweight edge calculation at the sensing nodes; transmitting the processed data through a low-power wide area network; the calculation center verifies and fuses the data; constructing a seawater parameter space-time distribution model; dynamically adjusting an energy consumption evaluation reference value by using a Pareto optimal solution set method; and periodically evaluating system energy consumption efficiency and updating model parameters. According to the method, dynamic balance of energy consumption and calculation efficiency of the seawater parameter monitoring system is realized, and the continuous operation capability and the monitoring reliability of the system are remarkably improved.

The invention relates to the technical field of calculation, and discloses a slag metal recovery process optimization method based on digital twinning, which comprises a data acquisition module, a data driving model construction module, a digital twinning construction module, a synchronous operation module, an online calibration module, a simulation optimization module, an online operation module and a model self-learning module. According to the method, through prospective simulation optimization, the recovery rate of each batch of materials is directly pushed to the theoretical limit; cost reduction and benefit increase are realized from the system level through a hybrid model and multi-objective optimization; through digital twin and automatic execution, process standardization and intelligent control are realized, and the stability and consistency of product quality are significantly improved. By adjusting model parameters and process constraints, smelteries with different scales and different raw material sources can be rapidly adapted, universality is high, standardization and productization are easy, and the method has high commercial expansion potential and huge market space.

The invention belongs to the technical field of nuclear reactor physical calculation, and particularly discloses a multi-group steady-state neutron transport coarse mesh node global efficient solving method and system, which effectively fuse and exert the advantages of a coarse mesh node method and a Newton-Kralov method. A general coarse mesh segment transport calculation model suitable for various polyhedral segments is constructed, high-precision and high-efficiency transport scanning calculation is realized under sparse segments, the number of variables is greatly reduced, and the calculation scale of single transport scanning is reduced; a Newton-Kralov subspace global solution framework is established, a global solution variable is selected based on a variable coupling relationship of a coarse network nodal method, a linear equation corresponding to a Newton step length is solved by using a Kralov subspace method, and the Newton step length is used for updating the global solution variable to realize rapid and efficient convergence. On the premise of ensuring the calculation precision, the calculation efficiency is remarkably improved, and the calculation scale is reduced, so that the method is particularly suitable for the problem of large-scale and high-dominance-ratio neutron transportation.

The invention belongs to the technical field of nuclear reactor physical numerical calculation, and particularly relates to a method for improving precision and convergence of solving a diffusion equation based on a PINN, and the method comprises the following steps: S1, building a neural network algorithm model combining a physical information neural network PINN and source iteration; s2, aiming at a multi-region smooth transition neutron diffusion problem with small cross section difference between materials, optimizing the neural network algorithm model by adopting an optimization strategy of a training level; and S3, aiming at the multi-region neutron diffusion problem of large neutron flux gradient at the boundary due to large cross section difference between materials, improving the neural network algorithm model by adopting a domain-divided neural network architecture. According to the method, the problem that the calculation cost is exponentially increased in a high-dimensional or complex geometric scene due to the fact that a traditional finite difference method and a finite element method depend on grid division is avoided.

The invention relates to the technical field of artificial intelligence and emotion calculation, and discloses an emotion state evaluation method based on multi-modal data, and the method comprises the steps: obtaining and decoupling an implicit signal, an explicit signal and situation data of an evaluated object; and respectively generating implicit features, explicit features and context vectors through double-flow decoupling coding and context knowledge graph coding. And the core engine determines the internal emotional state and the extrinsic emotional representation in parallel, and dynamically regulates and controls a bridging conversion process from the internal state to the predicted extrinsic representation by utilizing the situation vector. Finally, the emotion regulation intensity is calculated by comparing actual and predicted extrinsic representations, and the emotion regulation intensity, the internal state and the extrinsic representations are jointly used as a multi-dimensional evaluation result to be output. According to the method, the dynamic relationship and the adjustment strategy between the internal feeling and the external expression can be disclosed, a deeper and more three-dimensional analysis view angle is provided for the fields of mental health, human-computer interaction and the like, and the method has remarkable application value.

The embodiment of the invention provides a multi-agent collaborative scheduling method based on federal reinforcement learning and digital twinning, which belongs to the technical field of computing, and specifically comprises the following steps: importing a test task list, environmental parameters and an agent capability matrix; constructing a mathematical model of a scheduling decision; based on a digital twinborn model of a physical entity, rehearsing a scheduling process and outputting a pre-optimization strategy set containing a conflict avoidance strategy; independently executing a deep reinforcement learning algorithm locally, generating a private experience pool, and updating local model parameters according to the private experience pool; the aggregation server calculates an aggregation weight based on the performance index of each agent, and adopts a weighted average strategy to aggregate local model parameters of each agent; dynamically allocating tasks to corresponding agents and monitoring task execution states; real-time resource competition among the intelligent agents is solved through an evolutionary game mechanism; and performing scheduling ending evaluation and iteration. Through the scheme of the invention, the scheduling efficiency, security and robustness are improved.

The invention relates to the technical field of calculation, and provides a CFD solution acceleration method and system based on parallel calculation and load dynamic balance. Physical characteristics such as a CFL number and a residual error are quantified into load indexes, so that a load balancing decision and a flow field evolution rule are tightly coupled; by adopting an asynchronous communication overlapping framework deeply coupled with a CFD solver, blocked communication waiting time is converted into effective calculation time, and communication delay is systematically hidden; based on an incremental grid migration strategy of real-time load evaluation, high-load grid blocks in a hotspot process can be intelligently identified and migrated, and differentiation and high-efficiency utilization of computing power are realized.

The invention relates to the technical field of derivative variable calculation, and discloses a derivative variable calculation method and system based on a graph algorithm, and the method comprises the steps: analyzing derivative variable definition and calculation logic through employing a variable SQL syntax tree analysis technology, taking an analyzed variable as a node, determining a variable dependency relationship as an edge according to a processing task, and carrying out the calculation of the derivative variable; constructing a directed acyclic graph and storing the directed acyclic graph in a graph database; the nodes store variable metadata, execution priorities and calculation information, and store calculation task metadata and calculation information; according to node priorities, finding out high-priority variable nodes, querying all in-degree edges of the high-priority variable nodes, forming one or more subtrees, forming high-priority execution subtrees, performing topological sorting on the execution subtrees, and performing execution by an execution engine; the method aims to solve the problems of low calculation efficiency and insufficient utilization of calculation resources in the prior art.

The invention provides a generative intelligent optimization method and device, and relates to the technical field of generative artificial intelligence optimization computation.The method comprises the steps that a mathematical model containing decision variables, an objective function and constraint conditions is established for a target optimization problem, and scene parameters are pre-calculated through a traditional optimization algorithm to obtain a theoretical optimal solution; constructing a training data set containing scene parameters and a theoretical optimal solution; training the data set through a conditional generation type artificial intelligence model, minimizing the difference between a generation decision and a theoretical optimal solution, and establishing a mapping relation from scene parameters to an optimal decision, so that the mapping relation implicitly learns a constraint condition satisfaction mode; current scene parameters are collected in real time and input into the trained model, a near-optimal decision scheme is generated through reverse denoising or hidden variable decoding, and the near-optimal decision scheme is applied to real-time scenes such as industrial control. According to the method, a high-quality solution close to theoretical optimum is realized, the online solution time consumption is remarkably reduced, and the real-time requirement is met.

The invention relates to the technical field of cloud computing, and provides a cloud native assembly type application construction method and device supporting dynamic combination. According to the method, through component standardized packaging and distributed warehouse management, the high component reuse rate is achieved, the repeated development workload is reduced, a visual arrangement tool and a process template are reused, the application construction period is shortened by 50% or above, business personnel can directly participate in development, the'business-technology 'communication cost is reduced, and the development efficiency is improved. And during dynamic combination, an optimal loading path is calculated by adopting an optimization particle swarm algorithm of global search of a particle swarm algorithm and local kick of a simulated annealing algorithm, so that intelligent dependency analysis and optimal path calculation are realized, and the complexity of manual dependency sorting is avoided.

The invention relates to the technical field of bearing life calculation, in particular to a transmission bearing installation pre-tightening amount calculation method and system, and the method comprises the steps: S1, building a finite element mesh model of a transmission housing and a shaft through finite element simulation software based on predefined material attributes; the material attributes comprise elastic modulus, Poisson's ratio and thermal expansion coefficient of the transmission shell and the shaft; s2, based on the actual working condition of transmission assembly life simulation analysis, importing the finite element mesh model of the transmission shell and the shaft into a pre-constructed multi-body dynamics simulation model of the transmission shell, and establishing node connection between the finite element mesh model of the transmission shell and the shaft and a transmission part in the multi-body dynamics simulation model; s3, setting a detection object and a design variable, and adopting a preset strategy to obtain a target mapping curve of the detection object and the design variable; compared with the prior art, the optimal relation between the pre-tightening amount and the service life of the bearing can be quickly determined, so that the service life of the bearing is maximized.

The invention relates to the technical field of edge computing, and provides an edge data acquisition and processing device for a power plant unit, which comprises a data receiving module, an edge processing module, a data compression module and a communication module, the data receiving module is used for receiving data input by the access equipment and carrying out frame decoding processing on the data input by the access equipment to obtain data after frame decoding; the edge processing module is used for calculating and analyzing the deframed data to obtain preprocessed data; the data compression module is used for performing data compression processing on the preprocessed data to obtain compressed data; and the communication module is used for transmitting the compressed data to a preset target system. The efficient edge calculation and data compression mechanism not only improves the real-time performance and accuracy of data processing, but also enhances the adaptability of the system to the constantly changing operating environment and technical upgrading requirements of the power plant unit, and provides powerful support for safe operation and stability of the power plant unit.

The invention discloses a CFD (computational fluid dynamics) boundary processing acceleration method based on GPU (graphics processing unit) parallelism, and belongs to the technical field of fluid mechanics related calculation. The problem that in the prior art, a traditional CFD boundary processing acceleration method is low in calculation efficiency is solved. The method comprises the following steps: generating a computational grid required for simulating a flow field, and reading grid information and boundary condition information of computational simulation; initializing a flow field according to the calculation parameters, copying flow field data and flow field boundary information from a main memory to a video memory, establishing a boundary pointer array in the video memory, and instantiating a boundary pointed by each pointer; according to the boundary pointer array, the flow field data and the flow field boundary information, time step cycle calculation is carried out by utilizing a GPU, and physical quantity updating of a boundary virtual unit corresponding to the thread is realized; and after time step cycle calculation convergence, video memory flow field data are copied to a main memory, and flow field simulation calculation is completed. The method effectively improves the calculation efficiency, and can be applied to CFD boundary processing.

The invention relates to the technical field of confidential computing, and discloses a data credibility-oriented vehicle-mounted integrated security computing system and credibility construction method, which comprises the following steps of: establishing a static trust chain by using a hardware trust root, shielding external interruption in a credible execution environment, and controlling a performance monitoring unit; synchronously collecting real-time hardware micro-architecture events of the key business algorithm to generate a runtime feature vector; according to the method, the instruction stream during operation is anchored by utilizing the physical characteristics of the micro-architecture, the hijacking attack of the control stream is accurately identified in a ciphertext environment, the hijacking attack of the control stream is accurately identified, the hijacking attack of the control stream is accurately identified, the hijacking attack is accurately identified, and the hijacking attack of the control stream is accurately identified. Strong coupling verification of a calculation result and an execution behavior is realized, and logic safety of a vehicle-mounted platform is ensured.

The invention provides an energy storage comprehensive benefit evaluation method and system, and relates to the technical field of energy storage comprehensive benefit evaluation. Technical operation parameters, environmental parameters and market dynamic data of the energy storage system are collected in real time; automatically matching a preset evaluation mode through a scene recognition engine; calculating the comprehensive score of the technology-economy-society three-dimensional benefits by adopting a dynamic weighting algorithm to generate a calculation result; and generating an evaluation report containing the carbon footprint traceability code according to the calculation result. Through innovative technologies such as dynamic weight distribution, accurate compensation of battery attenuation and block chain data storage, the accuracy, adaptability and credibility of benefit evaluation of the energy storage system are remarkably improved.

The invention discloses a quantum cognitive simulation system and method based on classical computation, and belongs to the technical field of quantum computation.A data processing module receives original data of external data input equipment, converts the original data into quantum state data after processing, and transmits the quantum state data to a quantum cognitive simulation computation module; the quantum cognitive simulation calculation module performs quantum cognitive simulation calculation by using a CPU cluster, a GPU array and a coprocessor, and stores a result in the storage module or directly transmits the result to the output module; and the output module performs format conversion on the calculation result and outputs the result to external display / control equipment. The quantum cognitive simulation system and method based on classical calculation aim at overcoming the defects of a traditional cognitive model in the aspects of processing decision fuzziness, context dependency and high-dimensional calculation efficiency, the effects of quantum superposition, interference, entanglement and the like can be efficiently simulated, quantitative calculation of related functions of human consciousness is achieved, and the quantum cognitive simulation system and method based on classical calculation can be applied to the field of quantum computing. And the performance in an actual application scene is optimized.

The invention belongs to the technical field of calculation of horizontal bearing capacity of foundations, and particularly discloses and provides an accurate calculation method for horizontal bearing capacity of an offshore wind power foundation, which comprises the following steps: integrating wave spectrum, wind regime, tide and soil mass parameters to form a multi-dimensional data set; a three-dimensional pile-soil model is constructed, and wave force and wind load are coupled; calculating a stiffness coefficient and identifying a pile body type in combination with a length-diameter ratio; quantifying an environment deviation compensation coefficient according to the pile body type; after the initial horizontal bearing capacity is calculated, the bearing capacity failure critical state is judged based on real-time displacement monitoring data, and if the critical state is reached, the initial bearing capacity is corrected through linear weighting of the displacement deviation compensation coefficient and the environment deviation compensation coefficient. According to the method, the time-varying environment load is dynamically coupled, the pile-soil interaction is finely simulated, a rigidity self-adaptive grading mechanism and a double compensation feedback system are established, and the calculation precision and the real-time early warning capability of the horizontal bearing capacity in the complex marine environment are remarkably improved.

The embodiment disclosed by the application belongs to the technical field of computing, and particularly relates to a method and device for performing stencil calculation and a computing device. The method comprises the following steps: acquiring a plurality of node data included in a calculation region, and grouping the plurality of node data into a data vector. According to at least one calculation condition included in the stencil calculation, the plurality of node data are subjected to judgment processing, and a judgment result corresponding to each calculation condition is obtained for each node data. For each calculation condition, the judgment result corresponding to the calculation condition is grouped into a judgment result vector, and the judgment result vector is obtained. Based on the data vector and the judgment result vector corresponding to each calculation condition, a calculation operation corresponding to each calculation condition in the stencil calculation is performed, and a calculation result of the stencil calculation is obtained. The application can convert the stencil calculation into vector calculation, and can improve the efficiency of the stencil calculation.

The invention provides a watermark processing method and device, watermark processing equipment, a program product and a medium, and relates to the technical field of privacy computing. The method comprises the following steps: acquiring first data which is requested to be input into a target model by a user through an application programming interface (API); aiming at the first data, judging whether watermarking processing needs to be carried out or not; when it is determined that watermark processing needs to be carried out, based on a backdoor function corresponding to the target model, an API response corresponding to the API request is generated according to the first data, and the backdoor function is constructed according to a decision function of the target model; the API response is output to the user, the first data is added to a trigger set corresponding to the target model, and the trigger set is used for verifying an alternative model of the target model. According to the scheme of the invention, the problem that the existing digital watermarking technology is difficult to effectively defend model extraction attacks initiated through an API (Application Program Interface) is solved.

The invention provides a heterogeneous computing node optimization method and system for nuclear power computing, relates to the technical field of nuclear power computing, and constructs a heterogeneous computing node collaboration framework in a nuclear power computing scene, and the heterogeneous computing node collaboration framework comprises a node performance description layer, a task demand mapping layer and a resource scheduling execution layer. Firstly, calculation node feature collection is carried out on a node performance description layer to obtain a node feature set; establishing a matching rule base of a task module and a node feature set on a task demand mapping layer based on a nuclear power calculation task; then dynamic binding processing from the task module to the computing node is executed on the resource scheduling execution layer, and a task node binding relation table is generated; finally, data transmission path optimization processing between the nodes is executed based on the task node binding relation table and the communication link attributes, and a heterogeneous node optimization scheduling scheme is generated. According to the method, nuclear power calculation efficiency can be effectively improved, resource utilization is optimized, and data transmission delay is reduced.

The invention discloses a digital in-memory computing accelerator, an instruction set architecture and a data processing method thereof, and belongs to the technical field of in-memory computing. The digital memory computing accelerator comprises a chip-level module, a core-level module and a unit-level module, the chip-level module is used for acquiring a target calculation task, splitting the target calculation task into a plurality of core-level calculation tasks according to a load distribution strategy and distributing the core-level calculation tasks to the core-level module; the core-level module comprises a control unit, a calculation unit and a cache unit, and the control unit comprises an instruction fetching module, a decoding module and an instruction scheduler; an instruction fetching module reads a to-be-executed instruction corresponding to the core-level calculation task, a decoding module decodes the to-be-executed instruction to generate a control signal and sends the control signal to a calculation unit, and the to-be-executed instruction is an instruction of a calculation instruction set architecture in a digital memory; the calculation unit executes corresponding calculation operation according to the control signal; the cache unit caches intermediate data generated by executing the corresponding calculation operation.

The invention discloses an intelligent compilation method for a feasibility research report fusing a big data model and artificial intelligence, relates to the technical field of information, and solves the technical problem that it is difficult to obtain quantized values from multiple aspects of market scale, project income and technology to analyze project feasibility. Comprising the following steps: collecting and preprocessing multi-source data; an ARIMA model is constructed to output a market scale prediction value, and partition assignment is carried out; processing corresponding structured data through XGBoost, analyzing time series data through Transform, predicting ROI and CBR through a full connection layer after feature fusion, and weighting to generate a comprehensive income evaluation index prediction value; calculating a technical equipment performance evaluation index and a technical improvement evaluation index; constructing a linear regression model, and outputting a comprehensive technology evaluation index prediction value; and calculating a score according to the market scale partition value, the comprehensive income evaluation index prediction value and the comprehensive technical evaluation index prediction value, comparing the score with a threshold value, and automatically generating an analysis report.

The invention discloses a privacy computing arrangement method, system and device based on dynamic trust and a medium, and relates to the technical field of data security. The method comprises the following steps: receiving a data processing request, and collecting context data including user trust, environmental risk and data sensitivity; calculating a dynamic trust score based on the context data, and dynamically selecting a target privacy calculation technology according to the score; scheduling tasks based on the selected technology and deploying a guard agent in the computing container; running state data is periodically collected through a guard agent to serve as heartbeat for verification, and if verification fails, key interruption and memory erasing and fusing tasks are triggered; after calculation is completed, packaging a result and an access control strategy into a security capsule; and in response to the capsule opening request, decrypting the rendering result in the memory after verifying that the terminal environment conforms to the strategy. According to the method, full-life-cycle dynamic control from calculation request to result use is realized, and the problem that authorization is out of control is solved.

The invention relates to the technical field of multi-party security computing, and discloses a high-performance reliable multi-party security computing technology test method and device and a storage medium, and the method comprises the steps: deploying a to-be-tested multi-party security computing platform and a corresponding preset test data set in a preset test system simulating multi-party participation; simulating multi-party calculation to obtain a ciphertext calculation result, and comparing an error between the ciphertext calculation result and the plaintext calculation result to obtain a first reliability confirmation result; the ciphertext calculation is repeatedly executed for multiple times, and time consumed by each time of calculation is obtained; comparing each time of time consumption with a preset time threshold to obtain a second reliability confirmation result; comparing the performance difference between the ciphertext training model and the plaintext training model to obtain a third reliability confirmation result; and when the three reliability confirmation results all meet the corresponding reliability judgment standards, confirming that the to-be-tested multi-party security computing platform is reliable. Therefore, the stability and credibility of the MPC platform in the actual service can be evaluated more comprehensively and accurately.