Multi-mode-based industrial internet production scheduling method and system

Abstract

The invention relates to the technical field of industrial data processing, and provides an industrial internet production scheduling method and system based on multiple modes. Industrial data in the production process are obtained through an intelligent sensor in the industrial internet, modal parameters corresponding to all modals are calculated, and space-time alignment is carried out to generate an industrial tensor; decomposing and combining the industrial tensor through a preset first decomposition mode and a preset second decomposition mode to obtain a data tensor; generating a scheduling strategy based on to-be-optimized parameters screened from a parameter space of production scheduling and a preset Lie algebra base, and constructing a strategy function based on the scheduling strategy and the data tensor; and adjusting a modal weight factor in the strategy function to optimize a scheduling strategy in the strategy function to generate a target strategy, and performing distributed control on each industrial device through the industrial control terminal. The digital twin base constructed through the intelligent sensor ensures the real-time performance of data, the scheduling and production efficiency is improved, and the idle rate of equipment is reduced.

Description

Technical Field

[0001] The present application relates to the field of industrial data processing technology, and specifically, to a multimodal-based industrial Internet production scheduling method and system. Background Art

[0002] Driven by both Industry 4.0 and intelligent manufacturing, traditional production scheduling models are facing multiple challenges, including the expansion of device connectivity, explosive growth in data dimensions, and a surge in the demand for real-time response. Traditional Industrial Internet scheduling systems rely on single data sources, such as equipment operation logs and sensor time series data. For example, in semiconductor wafer manufacturing, relying solely on temperature sensor data cannot accurately predict thermal stress deformation. Multimodal fusion analysis combining optical inspection images and process parameters is required.

[0003] Currently, multimodal industrial scheduling still faces multiple challenges. For example, there are difficulties in processing heterogeneous data and technical barriers to semantic synchronization of data from different modalities, which can lead to distortions in information and feature fusion. This in turn makes it impossible to determine the current state of industrial production based on collected data, resulting in low efficiency and accuracy in production scheduling. Summary of the Invention

[0004] The present application provides a multimodal industrial Internet production scheduling method and system, which can at least to a certain extent solve the problem of low efficiency and accuracy of production scheduling.

[0005] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the present application.

[0006] According to one aspect of the present application, a multimodal industrial Internet production scheduling method is provided, comprising: acquiring industrial data in a production process through intelligent sensors in the industrial Internet; calculating modal parameters corresponding to each modality based on the multimodal industrial data, and performing spatiotemporal alignment of the industrial data through the modal parameters to generate an industrial tensor; decomposing the industrial tensor through a preset first decomposition method and a second decomposition method, respectively, and combining the generated first tensor and second tensor to obtain a data tensor; generating a scheduling strategy based on parameters to be optimized screened from a parameter space of production scheduling and a preset Lie algebraic basis, and constructing a strategy function based on the scheduling strategy and the data tensor; adjusting the modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function to generate a target strategy for real-time scheduling of industrial production; sending the target strategy to an industrial control terminal, and performing distributed control of each industrial equipment through the industrial control terminal.

[0007] In the present application, based on the aforementioned scheme, the multi-modal industrial data calculates the modal parameters corresponding to each mode, and the industrial data is spatially and temporally aligned through the modal parameters to generate an industrial tensor, including: determining the parameter vector and eigenvector of each mode based on the industrial data, and determining the modal parameters corresponding to each mode according to the parameter vector and eigenvector; performing a nonlinear transformation on the industrial data to generate a first result, and spatially and temporally aligning the first result through the modal parameters to generate an industrial tensor.

[0008] In the present application, based on the above-mentioned scheme, the parameter vector and eigenvector of each mode are determined based on the industrial data, and the modal parameters corresponding to each mode are determined according to the parameter vector and eigenvector, including: extracting the parameter vector and eigenvector of each mode from the industrial data; calculating the modal parameter ω corresponding to the kth mode based on the parameter vector and eigenvector of each mode k for:

[0009]

[0010] Among them, α k represents the parameter vector corresponding to the kth modal data, W represents the preset weight matrix, h k represents the feature vector extracted from the k-th modal data in the neural network, h k′ represents the feature vector extracted from the k′th modal data in the neural network, b represents the preset bias vector, and k′ represents the summation index.

[0011] In the present application, based on the above-mentioned solution, the nonlinear transformation of the industrial data is performed to generate a first result, and the first result is aligned in time and space by the modal parameters to generate an industrial tensor, including: based on the mode corresponding to the industrial data, the nonlinear transformation of the industrial data is performed to generate the first result as The first result is aligned in time and space by the modal parameters to generate the industrial tensor:

[0012]

[0013] Among them, X t，s，m，f Represents the industrial tensor in time t, space s, mode m and feature f, Φ k (·) represents the nonlinear transformation function of the kth mode, represents the industrial data of the mth mode at the (t, s) time-space point, K represents the total number of modes of industrial data, and f and F represent the identification and total number of features, respectively.

[0014] In the present application, based on the aforementioned scheme, the industrial tensor is decomposed by a preset first decomposition method and a second decomposition method respectively, and the generated first tensor and second tensor are combined to obtain a data tensor, including: decomposing the industrial tensor into a first tensor and multiple factor matrices through the first decomposition method; determining the characteristic parameters for decomposition through the first tensor and the factor matrix; decomposing the industrial tensor into multiple second tensors based on the characteristic parameters through the second decomposition method; and generating a data tensor based on the first tensor and the second tensor.

[0015] In the present application, based on the aforementioned scheme, the scheduling strategy is generated based on the parameters to be optimized screened from the parameter space of production scheduling and the preset Lie algebraic basis, and the strategy function is constructed based on the scheduling strategy and the data tensor, including: obtaining the parameter space for production scheduling, and determining the parameters to be optimized therefrom; generating the scheduling strategy through exponential mapping based on the parameters to be optimized and the preset Lie algebraic basis; and constructing the strategy function based on the scheduling strategy and the data tensor.

[0016] In the present application, based on the aforementioned scheme, the modal weight factor in the strategy function is adjusted to optimize the scheduling strategy in the strategy function and generate a target strategy for real-time scheduling of industrial production, including: adjusting the modal weight factor in the strategy function to obtain the minimum value of the strategy function; based on the minimum value of the strategy function, determining the scheduling strategy in the strategy function and generating a target strategy for real-time scheduling of industrial production.

[0017] In the present application, based on the aforementioned scheme, the target strategy is sent to the industrial control terminal, and distributed control of each industrial device is performed through the industrial control terminal, including: encoding the target strategy to generate control instructions; sending the control instructions to industrial control terminals distributed in different locations, and distributed control of each industrial device is performed through the industrial control terminal.

[0018] In this application, based on the aforementioned solution, the industrial data includes equipment data, order data, operation data, and logistics data.

[0019] According to one aspect of the present application, a multimodal industrial Internet production scheduling system is provided, comprising:

[0020] The acquisition unit is used to obtain industrial data in the production process through smart sensors in the industrial Internet;

[0021] an alignment unit, configured to calculate modal parameters corresponding to each mode based on multimodal industrial data, and perform spatiotemporal alignment on the industrial data using the modal parameters to generate an industrial tensor;

[0022] A tensor unit is configured to decompose the industrial tensor using a preset first decomposition method and a second decomposition method, respectively, and combine the generated first tensor and second tensor to obtain a data tensor;

[0023] A function unit, configured to generate a scheduling strategy based on parameters to be optimized screened from a parameter space of production scheduling and a preset Lie algebraic basis, and to construct a strategy function based on the scheduling strategy and the data tensor;

[0024] A strategy unit, configured to adjust a modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function and generate a target strategy for real-time scheduling of industrial production;

[0025] The control unit is used to send the target strategy to the industrial control terminal, and perform distributed control on each industrial device through the industrial control terminal.

[0026] According to one aspect of the present application, a computer-readable medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the multimodal industrial Internet production scheduling method as described in the above embodiment is implemented.

[0027] According to one aspect of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the multimodal industrial Internet production scheduling method as described in the above embodiments.

[0028] According to one aspect of the present application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the multimodal industrial Internet production scheduling method provided in the various optional implementations described above.

[0029] The technical solution of this application uses smart sensors in the industrial Internet to obtain industrial data from the production process; calculates the modal parameters corresponding to each mode based on multi-modal industrial data, and uses the modal parameters to align the industrial data in time and space to generate an industrial tensor; decomposes the industrial tensor using a preset first decomposition method and a preset second decomposition method, respectively, and combines the generated first tensor and second tensor to obtain a data tensor; generates a scheduling strategy based on the parameters to be optimized selected from the parameter space of production scheduling and a preset Lie algebra basis, and constructs a strategy function based on the scheduling strategy and the data tensor; adjusts the modal weight factors in the strategy function to optimize the scheduling strategy in the strategy function, generates a target strategy for real-time scheduling of industrial production, sends the target strategy to the industrial control terminal, and performs distributed control of various industrial equipment through the industrial control terminal. The digital twin base constructed by smart sensors ensures data real-time, and the spatiotemporal alignment and dual decomposition mechanism convert heterogeneous data into compact tensors. The adaptive adjustment based on multi-modality takes into account industrial constraints and dynamic optimization, reduces equipment idle rate, and improves the efficiency and accuracy of scheduling and production.

[0030] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification, are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and those skilled in the art can derive other drawings based on these drawings without inventive effort.

[0032] Figure 1 A flowchart of a multimodal industrial Internet production scheduling method in one embodiment of the present application is schematically shown.

[0033] Figure 2 The flowchart of generating a data tensor in one embodiment of the present application is schematically shown.

[0034] Figure 3 A schematic diagram of a multimodal industrial Internet production scheduling system in one embodiment of the present application is shown schematically.

[0035] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0036] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this application will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0037] In addition, described feature, structure or characteristic can be combined in one or more embodiments in any suitable manner.In the following description, many specific details are provided so as to provide a full understanding of the embodiments of the present application. However, it will be appreciated by those skilled in the art that the technical scheme of the present application can be put into practice without one or more of the specific details, or other methods, components, devices, steps etc. can be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the application.

[0038] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically separate entities. That is, these functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0039] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, while others may be combined or partially combined. Therefore, the actual execution order may vary depending on the actual situation.

[0040] The implementation details of the technical solution of this application are described in detail below:

[0041] Figure 1 The flowchart of the multimodal industrial Internet production scheduling method according to one embodiment of the present application is shown. Figure 1 As shown, the multimodal industrial Internet production scheduling method includes at least steps S110 to S160, which are described in detail as follows:

[0042] S110, obtaining industrial data in the production process through smart sensors in the industrial Internet, wherein the industrial data includes equipment data, order data, operation data and logistics data.

[0043] In one embodiment of the present application, multiple types of smart sensors are deployed at the production site. For example, vibration sensors, temperature sensors, and radio frequency identification readers are used to collect equipment data, such as operating parameters. Programmable logic controllers are used to obtain order data, such as order fulfillment status. Operational data, such as work order data, is synchronized through a production execution system interface. Furthermore, vehicle-mounted terminals and the global positioning system of automated guided vehicles (AGVs) are used to track logistics data, such as material flow paths, in real time.

[0044] Smart sensors use industrial Ethernet or wireless communication protocols to transmit raw data such as device operating parameters and material flow records to edge gateways in real time via industrial bus protocols. Edge computing nodes perform preliminary cleaning of industrial data, including outlier filtering and data format conversion, to ensure consistency across time and space for heterogeneous multi-source data.

[0045] At the data transmission level, a layered communication architecture is constructed. Field-layer devices achieve synchronous transmission via a time-sensitive network, ensuring the real-time delivery of control commands. The edge layer encapsulates pre-processed data into standard message bodies, passes deep packet inspection through the industrial firewall, and uploads them to the cloud message queue. To cope with massive amounts of concurrent data, a backpressure mechanism and consumer group technology are implemented to ensure reliable data transmission.

[0046] An industrial big data platform is established in the cloud or on a local server, storing preprocessed structured data by modality. Equipment data is stored in a time-series database to support high-frequency queries. Order and job data are linked to the production process using a graph database, and logistics data is dynamically tracked using spatial indexing technology. The platform also deploys a data quality monitoring module to continuously verify the integrity, consistency, and timeliness of data across all modalities, providing a reliable data foundation for subsequent tensor modeling.

[0047] S120 , calculating modal parameters corresponding to each mode based on the multimodal industrial data, performing spatiotemporal alignment on the industrial data using the modal parameters, and generating an industrial tensor.

[0048] In this embodiment, after the industrial data is acquired, the collected multimodal industrial data is cleaned and standardized. Optionally, the sensor noise in the equipment data is eliminated by Kalman filtering, the structured fields in the order data are extracted using natural language processing technology, and the positioning offset in the logistics data is corrected by the trajectory clustering algorithm. Subsequently, each modal data is input into a pre-trained deep learning model, the equipment data uses a one-dimensional convolutional network to extract time series features, the order data is associated through a graph attention network modeling process, and the logistics data uses a spatiotemporal convolutional network to capture dynamic trajectory features. These feature vectors are fused with modality-specific parameter vectors (such as equipment health index, order priority weight) to form a modal representation containing spatiotemporal semantics.

[0049] During the spatiotemporal alignment phase, a tensor alignment algorithm is used to achieve multimodal data fusion. First, a global spatiotemporal coordinate system is constructed based on the spatiotemporal references of each modality, such as the equipment operating cycle, order delivery window, and logistics transportation period. Feature maps are then spatially transformed using a deformable convolutional network, and time axis alignment is achieved using a dynamic time warping algorithm. Modal parameters serve as attention weights in this process, dynamically adjusting the contribution of each modal feature. Ultimately, the aligned features are encoded into a four-dimensional industrial tensor Z(t, s, m, f), where the spatiotemporal dimension (t, s) records the dynamic process, the modal dimension (m) distinguishes data types, and the feature dimension (f) stores semantic information, forming a unified data structure that supports subsequent tensor decomposition.

[0050] In one embodiment of the present application, modal parameters corresponding to each mode are calculated based on multimodal industrial data, and the industrial data are spatiotemporally aligned using the modal parameters to generate an industrial tensor, including:

[0051] Determining a parameter vector and a eigenvector of each mode based on the industrial data, and determining a modal parameter corresponding to each mode according to the parameter vector and the eigenvector;

[0052] A nonlinear transformation is performed on the industrial data to generate a first result, and the first result is spatially and temporally aligned using the modal parameters to generate an industrial tensor.

[0053] In this embodiment, multimodal, heterogeneous data is preprocessed for spatiotemporal alignment and standardization. The input data includes equipment sensor data, production order text, worker behavior video streams, and supply chain logistics trajectories. These data come from different sources with varying formats and spatiotemporal references, necessitating preprocessing.

[0054] Parameter vectors and eigenvectors for each modality are extracted from the industrial data. Specifically, the parameter vector corresponding to the kth modal data includes a vector composed of parameters such as statistical characteristics related to the kth modal data and a preset importance index. The eigenvector represents a vector extracted from the industrial data using a neural network.

[0055] Based on the parameter vector and eigenvector of each mode, calculate the corresponding modal parameter ω of the kth mode k for:

[0056]

[0057] Among them, α k represents the parameter vector corresponding to the kth modal data; W represents the preset weight matrix, h k represents the feature vector extracted from the k-th modal data in the neural network, h k′represents the feature vector extracted from the k′th modal data in the neural network; b represents the preset bias vector, and k′ represents the summation index used to traverse all modalities. The modal parameters calculated in this embodiment are used to measure the importance of each modal industrial data in the fusion process.

[0058] By calculating the modal parameters corresponding to each mode, the data of different modes are aligned in time and space to generate the industrial tensor X t，s，m，f for:

[0059]

[0060] Among them, X t，s，m，f represents the industrial tensor in time t, space s, mode m and feature f, Φk(·) represents the nonlinear transformation function of the kth mode, represents the industrial data of the mth mode at the (t, s) time-space point, K represents the total number of modes, f and F represent the identity and total number of features, respectively.

[0061] In the above calculation process, the numerator retains multimodal complementary information through weighted fusion, that is, data from different modalities are fused according to their importance (determined by the attention weight); the denominator is normalized to eliminate dimensional differences and ensure that data from different modalities are comparable when fused.

[0062] This process solves the problem of inconsistent spatiotemporal benchmarks for heterogeneous data from multiple sources. It aligns data from different sources and types in both time and space, generating a spatiotemporally aligned four-dimensional tensor. This provides a unified representation framework for subsequent data fusion, eliminating dimensional differences between modal data and preserving the complementary information within multimodal data, facilitating further data analysis and processing.

[0063] S1 30, decomposing the industrial tensor by a preset first decomposition method and a preset second decomposition method respectively, combining the generated first tensor and second tensor to obtain a data tensor.

[0064] In an embodiment of the present application, a two-stage decomposition strategy can be used to perform structured analysis of industrial tensors. First, the high-dimensional tensor is decomposed into a combination of multiple core tensors and factor matrices through the first decomposition method. This process is similar to decomposing complex data into interpretable independent components, effectively eliminating redundant information and retaining key characteristic patterns. Subsequently, the second decomposition method further refines the factor matrix twice, and strengthens the physical interpretability by constraining the non-negativity of the decomposition results, ensuring that each component corresponds to an actual production factor, such as equipment load and order priority.

[0065] The two decomposition results are dynamically weighted and fused to generate a data tensor, which stitches together the observations from different perspectives. Optionally, the decomposition weights are automatically adjusted through a cross-validation mechanism, so that the data tensor retains the spatiotemporal continuity of the original data while highlighting the hierarchical structure of key scheduling indicators.

[0066] like Figure 2 As shown, in one embodiment of the present application, the industrial tensor is decomposed by a preset first decomposition method and a second decomposition method respectively, and the generated first tensor and second tensor are combined to obtain a data tensor, including:

[0067] S2 1 0, decomposing the industrial tensor into a first tensor and a plurality of factor matrices by a first decomposition method;

[0068] S220, determining characteristic parameters for decomposition using the first tensor and the factor matrix;

[0069] S230, decomposing the industrial tensor into a plurality of second tensors based on the characteristic parameters using a second decomposition method;

[0070] S240: Generate a data tensor based on the first tensor and the second tensor.

[0071] In one embodiment of the present application, the industrial tensor is decomposed into a first tensor and multiple factor matrices through a first decomposition method. For example, the input industrial tensor X is decomposed into a first tensor G and the product of three factor matrices A, B, and C, that is, X≈G×A×B×C. Among them, the first tensor G captures the interaction information between different modalities, and the factor matrices A, B, and C extract the principal components in the time, space, and modal dimensions respectively. The industrial tensor is decomposed into the product of the first tensor and each modal factor matrix through the first decomposition method to mine the high-order interaction features between multimodal data, retain the main interaction features, and improve the feature expression effect.

[0072] In one embodiment of the present application, the characteristic parameter λ for decomposition is determined by using the first tensor and the factor matrix. r for:

[0073]

[0074] Among them, a represents the preset weight, G (n) Represents the nth tensor slice in the first tensor, A (n) 、B (n) 、C (n) represents the nth matrix slice in the factor matrix, Tr(·) represents the trace of the matrix, ∑ r′ It represents the sum of the parameters of all tensors r′, where r′ represents the tensor identifier in the summation operation.

[0075] In one embodiment of the present application, the industrial tensor is decomposed into multiple second tensors based on the calculated characteristic parameters by a second decomposition method. Specifically, the input tensor X is decomposed into R second tensors of rank 1, that is:

[0076]

[0077] Among them, λ r is the weight coefficient, a r 、b r 、c r d r are the second vectors on each mode (time t, space s, mode m, and feature f), r and R represent the identity and total number of the second tensor, respectively. The second decomposition method outputs a low-rank representation tensor, providing a compact and information-rich feature input for subsequent scheduling strategy optimization.

[0078] In one embodiment of the present application, based on the first tensor and the second tensor, a data tensor Z is generated as follows:

[0079]

[0080] For example, this embodiment performs deep feature extraction and dimensionality compression on standardized industrial tensors. The first decomposition method captures the complex interactions between the tensor's modes, while the second decomposition method achieves a concise low-dimensional representation through tensor superposition. High-order tensor decomposition and a hybrid decomposition architecture enable high-order interaction feature mining and efficient dimensionality compression between multimodal data, achieving efficient feature extraction and providing strong support for subsequent scheduling strategy optimization.

[0081] S140 , generating a scheduling strategy based on the parameters to be optimized screened from the parameter space of production scheduling and a preset Lie algebraic basis, and constructing a strategy function based on the scheduling strategy and the data tensor.

[0082] Feature selection algorithms (such as random forest importance evaluation) are used to filter out optimization parameters from the parameter space that significantly impact scheduling objectives, such as equipment start / stop thresholds and order consolidation rule weights. These parameters are then mapped to a manifold space composed of Lie algebra basis vectors. Exponential mapping is then used to generate continuous scheduling trajectories that adhere to industrial constraints, ensuring that operations such as the robot's motion path and material handling sequence meet spatial continuity and dynamic feasibility.

[0083] The generated scheduling strategy is then fused with the preprocessed data tensor, which contains multi-dimensional information such as equipment status, order urgency, and worker efficiency, aligned in time and space. An attention mechanism dynamically assigns weights to each modality, constructing a policy function that integrates physical rules and data-driven decision-making.

[0084] In one embodiment of the present application, a scheduling strategy is generated based on parameters to be optimized screened from a parameter space of production scheduling and a preset Lie algebraic basis, and a strategy function is constructed based on the scheduling strategy and the data tensor, including:

[0085] Obtain the parameter space for production scheduling and determine the parameters to be optimized;

[0086] Based on the parameters to be optimized and a preset Lie algebraic basis, generating a scheduling strategy through exponential mapping;

[0087] A policy function is constructed based on the scheduling policy and the data tensor.

[0088] In one embodiment of the present application, a parameter space for production scheduling is obtained, data in the parameter space is detected based on a preset parameter threshold, and parameters exceeding the parameter threshold are screened as parameters to be optimized.

[0089] In one embodiment of the present application, based on the parameters to be optimized and the preset Lie algebraic base, a scheduling strategy is generated by exponential mapping. Specifically, the exponential mapping of the Lie group is used to combine the parameters to be optimized with the Lie algebraic base, and the scheduling strategy g is generated by exponential mapping as follows:

[0090]

[0091] Among them, θ c represents the parameters to be optimized, E c represents the Lie algebra basis corresponding to the parameter to be optimized, and c represents the data identifier of the parameter to be optimized.

[0092] Optionally, by adjusting the parameters to be optimized, different scheduling strategies g can be generated, and these scheduling strategies correspond to different production scheduling solutions.

[0093] In this process, Lie groups, as special Euclidean groups, are well-suited to representing rigid-body motion in industrial production, including translations and rotations. Through exponential mapping, the parameters to be optimized in Lie algebra space can be mapped to Lie group space, thereby deriving a specific scheduling strategy.

[0094] In one embodiment of the present application, based on the scheduling strategy and the data tensor, a strategy function J(g) is constructed as follows:

[0095]

[0096] in, represents the cumulative difference between the scheduling strategy and the actual demand in time [0, T], g(t) represents the scheduling strategy at time t, Z(t) represents the data tensor at time t, and d(t) represents the production demand tensor at time t. represents the square of the weighted norm, Q represents the preset modal weighting matrix, β represents the modal weight factor of the collaborative term, represents the summation operation of all indices satisfying i<j, Tr(·) represents the trace of the matrix, V represents the graph Laplacian matrix based on the process priority, || g i || F and ‖g j ‖ F Represents the scheduling strategy g i 、g j The norm of , i and j represent the identifiers of the scheduling strategy.

[0097] Optionally, by adjusting the modal weight factor β, the weight between the tracking error and the synergy term can be balanced to obtain a scheduling strategy that meets different production needs.

[0098] In this embodiment, the policy function consists of two parts: a tracking error term, which measures the discrepancy between the scheduling policy and actual demand; and a Lie group coordination term, which promotes coordination between different processes. By minimizing the policy function J(g), the constrained optimization problem is solved, enabling real-time dynamic adjustments. Using a variational inference framework and a dynamic weight adjustment mechanism, the scheduling policy is optimized and adapted to varying production conditions and demand changes.

[0099] By finding the policy parameters of the scheduling policy in the parameter space, the optimal scheduling policy g can be found, so that the scheduling policy can not only meet the production needs but also achieve collaboration between processes.

[0100] The above process transforms the discrete scheduling problem into a continuous manifold optimization problem, making the optimization process smoother and more efficient. Traditional discrete scheduling optimization processes can suffer from discontinuities and difficulty converging. However, continuous manifold optimization can leverage optimization algorithms like gradient descent to quickly find the optimal solution. The multi-objective game cost function comprehensively considers tracking error and process coordination, enabling multi-objective optimization.

[0101] S150, adjusting the modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function and generate a target strategy for real-time scheduling of industrial production.

[0102] In one embodiment of the present application, adjusting the modal weight factor in the policy function to optimize the scheduling policy in the policy function and generate a target policy for real-time scheduling of industrial production includes:

[0103] Adjusting the modal weight factor in the strategy function to obtain a minimum value of the strategy function;

[0104] Based on the minimum value of the policy function, a scheduling policy in the policy function is determined, and a target policy for real-time scheduling of industrial production is generated.

[0105] In one embodiment of the present application, the modal weight factor is dynamically adjusted based on the gradient direction of the objective function with respect to the modal weight factor and the system's busyness, achieving real-time dynamic adjustment. Simultaneously, a variational inference framework is utilized to optimize the scheduling policy by minimizing an objective function that includes the difference between the variational distribution and the true distribution, as well as a regularization term. Ultimately, the optimized target policy is output.

[0106] In this embodiment, the variational inference framework, based on variational inference methods, solves constrained non-convex optimization problems by approximating complex probability distributions. The variational distribution is used to approximate the true posterior distribution, and the optimized target policy and modal weight factors are obtained by minimizing the difference between the variational distribution and the true distribution. A dynamic weight adjustment mechanism adaptively adjusts the contribution of each modality based on the gradient direction and system activity level, ensuring the system maintains good adaptability and performance under different conditions.

[0107] The variational inference framework enables more efficient solutions to constrained non-convex optimization problems, improving the efficiency of solving optimization problems. Furthermore, the dynamic weight adjustment mechanism enables the system to dynamically adjust to real-time production conditions and demand changes, enhancing its real-time and dynamic adaptability. Ultimately, the optimized scheduling strategy is more stable and accurate, better meeting production needs and improving overall production efficiency and scheduling performance.

[0108] S160: Send the target strategy to an industrial control terminal, and perform distributed control on each industrial device through the industrial control terminal.

[0109] In one embodiment of the present application, the target policy is sent to an industrial control terminal, and distributed control of each industrial device is performed through the industrial control terminal, including:

[0110] Encoding the target strategy to generate control instructions;

[0111] The control instructions are sent to industrial control terminals distributed at different locations, and distributed control is performed on various industrial devices through the industrial control terminals.

[0112] In one embodiment of the present application, the target scheduling strategy is encoded into a control instruction, wherein the control instruction includes information such as the operation to be performed by each industrial device, parameter settings, and execution sequence.

[0113] Control commands are then sent to industrial control terminals distributed across various locations via a network or other communication methods. Each industrial control terminal is responsible for managing a group of industrial equipment within its area. Upon receiving the commands, the control terminal interprets them and, based on the results, precisely controls the corresponding industrial equipment, such as adjusting equipment operating status and changing production parameters. This ensures that the entire production process proceeds in an orderly manner according to the optimized scheduling strategy.

[0114] In this embodiment, the target strategy is sent to the control terminal, and the local computing and decision-making capabilities of the control terminal are utilized to achieve rapid and accurate control of industrial equipment.

[0115] Alternatively, complex control tasks can be broken down into multiple subtasks and assigned to multiple industrial control terminals for parallel execution. Each control terminal focuses on controlling the equipment within its area, and through local decision-making and collaborative work, they jointly achieve global production goals.

[0116] At the same time, the distributed control architecture improves the reliability and flexibility of the system. When a control terminal or device fails, other control terminals can take over its tasks to ensure the continuity and stability of the production process.

[0117] By sending target strategies to control terminals for distributed control, control efficiency and response speed can be significantly improved. Control terminals can receive and execute control commands in real time, rapidly adjusting equipment status to adapt to changes in the production process. Distributed control also reduces the burden on the central control system and improves overall system stability and reliability. Furthermore, this control approach facilitates system expansion and maintenance. When new industrial equipment or control terminals are added, they can be easily integrated into the existing system without requiring a major overhaul of the entire system.

[0118] The technical solution of this application uses smart sensors in the industrial Internet to obtain industrial data in the production process; calculates the modal parameters corresponding to each mode based on multi-modal industrial data, and uses the modal parameters to align the industrial data in time and space to generate an industrial tensor; decomposes the industrial tensor using a preset first decomposition method and a preset second decomposition method, respectively, and combines the generated first tensor and second tensor to obtain a data tensor; generates a scheduling strategy based on the parameters to be optimized selected from the parameter space of production scheduling and a preset Lie algebra basis, and constructs a strategy function based on the scheduling strategy and the data tensor; adjusts the modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function, generates a target strategy for real-time scheduling of industrial production, sends the target strategy to the industrial control terminal, and performs distributed control of various industrial equipment through the industrial control terminal. The digital twin base constructed by smart sensors ensures data real-time, and the spatiotemporal alignment and dual decomposition mechanism convert heterogeneous data into compact tensors. The adaptive adjustment based on multi-modality takes into account both industrial constraints and dynamic optimization, improves scheduling and production efficiency, and reduces equipment idle rate.

[0119] The following introduces an embodiment of the multimodal industrial Internet production scheduling system of the present application, which can be used to execute the multimodal industrial Internet production scheduling method in the above-mentioned embodiment of the present application. It can be understood that the multimodal industrial Internet production scheduling system can be a computer program (including program code) running on a computer device, for example, the multimodal industrial Internet production scheduling system is an application software; the multimodal industrial Internet production scheduling system can be used to execute the corresponding steps in the method provided in the embodiment of the present application. For details not disclosed in the embodiment of the multimodal industrial Internet production scheduling system of the present application, please refer to the embodiment of the multimodal industrial Internet production scheduling method mentioned above in the present application.

[0120] Figure 3 A block diagram of a multimodal industrial Internet production scheduling system according to an embodiment of the present application is shown.

[0121] Reference Figure 3 As shown, according to an embodiment of the present application, a multimodal industrial Internet production scheduling system includes:

[0122] An acquisition unit 310 is used to acquire industrial data in the production process through smart sensors in the industrial Internet;

[0123] an alignment unit 320 for calculating modal parameters corresponding to each mode based on the multimodal industrial data, and performing spatiotemporal alignment on the industrial data using the modal parameters to generate an industrial tensor;

[0124] A tensor unit 330 is configured to decompose the industrial tensor using a preset first decomposition method and a preset second decomposition method, and combine the generated first tensor and second tensor to obtain a data tensor;

[0125] A function unit 340 is configured to generate a scheduling strategy based on the parameters to be optimized selected from the parameter space of production scheduling and a preset Lie algebraic basis, and to construct a strategy function based on the scheduling strategy and the data tensor;

[0126] A strategy unit 350 is configured to adjust a modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function and generate a target strategy for real-time scheduling of industrial production;

[0127] The control unit 360 is used to send the target strategy to the industrial control terminal, and perform distributed control on each industrial device through the industrial control terminal.

[0128] In the present application, based on the aforementioned scheme, the multi-modal industrial data calculates the modal parameters corresponding to each mode, and the industrial data is spatially and temporally aligned through the modal parameters to generate an industrial tensor, including: determining the parameter vector and eigenvector of each mode based on the industrial data, and determining the modal parameters corresponding to each mode according to the parameter vector and eigenvector; performing a nonlinear transformation on the industrial data to generate a first result, and spatially and temporally aligning the first result through the modal parameters to generate an industrial tensor.

[0129] In the present application, based on the above scheme, the parameter vector and eigenvector of each mode are determined based on the industrial data, and the modal parameters corresponding to each mode are determined according to the parameter vector and eigenvector, including: calculating the modal parameter ω corresponding to the kth mode based on the parameter vector and eigenvector of each mode k for:

[0130]

[0131] Among them, α k represents the parameter vector corresponding to the kth modal data, W represents the preset weight matrix, h k represents the feature vector extracted from the k-th modal data in the neural network, h k′ represents the feature vector extracted from the k′th modal data in the neural network, b represents the preset bias vector, and k′ represents the summation index.

[0132] In the present application, based on the above-mentioned solution, the nonlinear transformation of the industrial data is performed to generate a first result, and the first result is aligned in time and space by the modal parameters to generate an industrial tensor, including: based on the mode corresponding to the industrial data, the nonlinear transformation of the industrial data is performed to generate the first result as The first result is aligned in time and space by the modal parameters to generate the industrial tensor:

[0133]

[0134] Among them, X t，s，m，f Represents the industrial tensor in time t, space s, mode m and feature f, Φ k (·) represents the nonlinear transformation function of the kth mode, represents the industrial data of the mth mode at the (t, s) time-space point, K represents the total number of modes of industrial data, and f and F represent the identification and total number of features, respectively.

[0135] In the present application, based on the aforementioned scheme, the industrial tensor is decomposed by a preset first decomposition method and a second decomposition method respectively, and the generated first tensor and second tensor are combined to obtain a data tensor, including: decomposing the industrial tensor into a first tensor and multiple factor matrices through the first decomposition method; determining the characteristic parameters for decomposition through the first tensor and the factor matrix; decomposing the industrial tensor into multiple second tensors based on the characteristic parameters through the second decomposition method; and generating a data tensor based on the first tensor and the second tensor.

[0136] In the present application, based on the aforementioned scheme, the scheduling strategy is generated based on the parameters to be optimized screened from the parameter space of production scheduling and the preset Lie algebraic basis, and the strategy function is constructed based on the scheduling strategy and the data tensor, including: obtaining the parameter space for production scheduling, and determining the parameters to be optimized therefrom; generating the scheduling strategy through exponential mapping based on the parameters to be optimized and the preset Lie algebraic basis; and constructing the strategy function based on the scheduling strategy and the data tensor.

[0137] In the present application, based on the aforementioned scheme, the modal weight factor in the strategy function is adjusted to optimize the scheduling strategy in the strategy function and generate a target strategy for real-time scheduling of industrial production, including: adjusting the modal weight factor in the strategy function to obtain the minimum value of the strategy function; based on the minimum value of the strategy function, determining the scheduling strategy in the strategy function and generating a target strategy for real-time scheduling of industrial production.

[0138] In the present application, based on the aforementioned scheme, the target strategy is sent to the industrial control terminal, and distributed control of each industrial device is performed through the industrial control terminal, including: encoding the target strategy to generate control instructions; sending the control instructions to industrial control terminals distributed in different locations, and distributed control of each industrial device is performed through the industrial control terminal.

[0139] In this application, based on the aforementioned solution, the industrial data includes equipment data, order data, operation data, and logistics data.

[0140] The technical solution of this application uses smart sensors in the industrial Internet to obtain industrial data in the production process; calculates the modal parameters corresponding to each mode based on multi-modal industrial data, and uses the modal parameters to align the industrial data in time and space to generate an industrial tensor; decomposes the industrial tensor using a preset first decomposition method and a preset second decomposition method, respectively, and combines the generated first tensor and second tensor to obtain a data tensor; generates a scheduling strategy based on the parameters to be optimized selected from the parameter space of production scheduling and a preset Lie algebra basis, and constructs a strategy function based on the scheduling strategy and the data tensor; adjusts the modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function, generates a target strategy for real-time scheduling of industrial production, sends the target strategy to the industrial control terminal, and performs distributed control of various industrial equipment through the industrial control terminal. The digital twin base constructed by smart sensors ensures data real-time, and the spatiotemporal alignment and dual decomposition mechanism convert heterogeneous data into compact tensors. The adaptive adjustment based on multi-modality takes into account both industrial constraints and dynamic optimization, improves scheduling and production efficiency, and reduces equipment idle rate.

[0141] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing an electronic device according to an embodiment of the present application is shown.

[0142] It should be noted that the computer system of the electronic device in this embodiment is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0143] In this embodiment, the computer system includes a central processing unit 401, which can perform various appropriate actions and processes based on programs stored in a read-only memory 402 or programs loaded from a storage unit 408 into a random access memory 403, such as executing the multimodal industrial Internet production scheduling method described in the above embodiment. The random access memory 403 also stores various programs and data required for system operation. The central processing unit 401, the read-only memory 402, and the random access memory 403 are connected to each other via a bus 404. An input / output interface 405 is also connected to the bus 404.

[0144] The following components are connected to the input / output interface 405: an input section 406 including a keyboard, a mouse, and the like; an output section 407 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 408 including a hard disk and the like; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card or a modem. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as needed. Removable media 411, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 410 as needed, so that computer programs read therefrom can be installed into the storage section 408 as needed.

[0145] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product that includes a computer program carried on a computer-readable medium, the computer program including a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 409 and / or installed from a removable medium 411. When the computer program is executed by the central processing unit 401, the various functions defined in the system of the present application are performed.

[0146] It should be noted that the computer-readable medium shown in the embodiments of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable computer program. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0147] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. Among them, each box in the flowchart or block diagram can represent a module, program segment, or part of the code, and the above-mentioned module, program segment, or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0148] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. In some cases, the names of these units do not constitute limitations on the units themselves.

[0149] According to one aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations described above.

[0150] As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or may exist independently without being incorporated into the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device implements the multimodal industrial Internet production scheduling method described in the above embodiments.

[0151] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiment of the application, the features and functions of two or more modules or units described above can be concretized in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0152] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the embodiments of the present application.

[0153] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of this application and include common knowledge or customary techniques in the art that are not disclosed herein.

[0154] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A multimodal industrial Internet production scheduling method, characterized in that: include: Obtain industrial data during the production process through smart sensors in the Industrial Internet; Calculating modal parameters corresponding to each mode based on multimodal industrial data, performing spatiotemporal alignment on the industrial data using the modal parameters, and generating an industrial tensor; Decomposing the industrial tensor by a preset first decomposition method and a second decomposition method respectively, and combining the generated first tensor and second tensor to obtain a data tensor; Generate a scheduling strategy based on parameters to be optimized screened from a parameter space of production scheduling and a preset Lie algebraic basis, and construct a strategy function based on the scheduling strategy and the data tensor; Adjusting the modal weight factor in the policy function to optimize the scheduling policy in the policy function and generate a target policy for real-time scheduling of industrial production; The target strategy is sent to an industrial control terminal, and distributed control is performed on each industrial device through the industrial control terminal.

2. The multimodal industrial Internet production scheduling method according to claim 1 is characterized in that: Calculating modal parameters corresponding to each mode based on multimodal industrial data, performing spatiotemporal alignment on the industrial data using the modal parameters, and generating an industrial tensor, including: Determining a parameter vector and a eigenvector of each mode based on the industrial data, and determining a modal parameter corresponding to each mode according to the parameter vector and the eigenvector; A nonlinear transformation is performed on the industrial data to generate a first result, and the first result is spatially and temporally aligned using the modal parameters to generate an industrial tensor.

3. The multimodal industrial Internet production scheduling method according to claim 2 is characterized in that: Determining a parameter vector and a eigenvector of each mode based on the industrial data, and determining a modal parameter corresponding to each mode according to the parameter vector and the eigenvector, including: Extracting parameter vectors and eigenvectors of each mode from the industrial data; Based on the parameter vector and eigenvector of each mode, calculate the corresponding modal parameter ω of the kth mode k for: Among them, α k represents the parameter vector corresponding to the kth modal data, W represents the preset weight matrix, h k represents the feature vector extracted from the k-th modal data in the neural network, h k′ represents the feature vector extracted from the k′th modal data in the neural network, b represents the preset bias vector, and k′ represents the summation index.

4. The multimodal industrial Internet production scheduling method according to claim 3 is characterized in that: Performing a nonlinear transformation on the industrial data to generate a first result, and performing spatiotemporal alignment on the first result using the modal parameters to generate an industrial tensor, including: Based on the mode corresponding to the industrial data, a nonlinear transformation is performed on the industrial data to generate a first result: The first result is aligned in time and space by the modal parameters to generate the industrial tensor: Among them, X t，s，m，f represents the industrial tensor in time t, space s, mode m and feature f, Φk(·) represents the nonlinear transformation of the industrial data of the kth mode, represents the industrial data of the mth mode at the (t, s) time-space point, K represents the total number of modes of industrial data, f and F represent the identification and total number of features respectively.

5. The multimodal industrial Internet production scheduling method according to claim 1, characterized in that: Decomposing the industrial tensor by a preset first decomposition method and a second decomposition method respectively, combining the generated first tensor and the second tensor to obtain a data tensor, including: Decomposing the industrial tensor into a first tensor and a plurality of factor matrices by a first decomposition method; Determining characteristic parameters for decomposition using the first tensor and the factor matrix; Decomposing the industrial tensor into a plurality of second tensors based on the characteristic parameters by a second decomposition method; A data tensor is generated based on the first tensor and the second tensor.

6. The multimodal industrial Internet production scheduling method according to claim 1, characterized in that: A scheduling strategy is generated based on parameters to be optimized screened from the parameter space of production scheduling and a preset Lie algebraic basis, and a strategy function is constructed based on the scheduling strategy and the data tensor, including: Obtain the parameter space for production scheduling and determine the parameters to be optimized; Based on the parameters to be optimized and a preset Lie algebraic basis, generating a scheduling strategy through exponential mapping; A policy function is constructed based on the scheduling policy and the data tensor.

7. The multimodal industrial Internet production scheduling method according to claim 1, characterized in that: Adjusting the modal weight factor in the policy function to optimize the scheduling policy in the policy function and generate a target policy for real-time scheduling of industrial production, including: Adjusting the modal weight factor in the strategy function to obtain a minimum value of the strategy function; Based on the minimum value of the policy function, a scheduling policy in the policy function is determined, and a target policy for real-time scheduling of industrial production is generated.

8. The multimodal industrial Internet production scheduling method according to claim 1, characterized in that: The target strategy is sent to an industrial control terminal, and distributed control of each industrial device is performed through the industrial control terminal, including: Encoding the target strategy to generate control instructions; The control instructions are sent to industrial control terminals distributed at different locations, and distributed control is performed on various industrial devices through the industrial control terminals.

9. The multimodal industrial Internet production scheduling method according to any one of claims 1 to 8, characterized in that: The industrial data includes equipment data, order data, operation data and logistics data.

10. A multimodal industrial Internet production scheduling system, characterized in that: include: The acquisition unit is used to obtain industrial data in the production process through smart sensors in the industrial Internet; an alignment unit, configured to calculate modal parameters corresponding to each mode based on multimodal industrial data, and perform spatiotemporal alignment on the industrial data using the modal parameters to generate an industrial tensor; A tensor unit is configured to decompose the industrial tensor using a preset first decomposition method and a second decomposition method, respectively, and combine the generated first tensor and second tensor to obtain a data tensor; A function unit, configured to generate a scheduling strategy based on parameters to be optimized screened from a parameter space of production scheduling and a preset Lie algebraic basis, and to construct a strategy function based on the scheduling strategy and the data tensor; A strategy unit, configured to adjust a modal weight factor in the strategy function to optimize the scheduling strategy in the strategy function and generate a target strategy for real-time scheduling of industrial production; The control unit is used to send the target strategy to the industrial control terminal, and perform distributed control on each industrial device through the industrial control terminal.

Images