Concrete production management method and system based on digital twinning

Abstract

The invention provides a concrete production management method and system based on digital twinning, and relates to the technical field of digital twinning, and the method comprises the steps: collecting the health monitoring data of concrete production equipment in real time through a plurality of sensors, and collecting the full life cycle event data of the concrete production equipment; preprocessing the health monitoring data at an edge end, extracting state features of the equipment, performing real-time diagnosis by using a pre-trained fault prediction model based on the state features of the equipment, and constructing an equipment multi-level digital twinborn model at a cloud platform; the method comprises the following steps: fusing a pre-constructed physical degradation model and a data driving model based on a multi-level digital twinborn model, constructing an equipment residual life prediction model, and dynamically adjusting a concrete production strategy in the digital twinborn model according to an obtained prediction result to obtain an optimal concrete production scheduling scheme. And the information is converted into an equipment control instruction, so that automatic issuing and execution of a concrete production scheduling scheme are realized.

Description

Technical Field

[0001] The present invention relates to digital twin technology, and in particular to a concrete production management method and system based on digital twin. Background Art

[0002] As the most important basic material in modern construction projects, the level of concrete production management directly affects the quality, progress and cost of the project. However, traditional concrete production management mainly relies on manual experience and simple information technology, which has problems such as insufficient perception of production status, difficulty in process optimization, and low efficiency in resource allocation, making it difficult to adapt to the increasingly complex needs of engineering construction.

[0003] In recent years, the rise of digital twin technology has provided new ideas for the intelligent transformation of concrete production management. Digital twins have been widely used in product design, manufacturing, operation and maintenance by building virtual mapping models of physical entities to achieve real-time interaction and integration between the physical world and digital space. The introduction of digital twins into concrete production management is expected to break through the limitations of traditional methods and achieve full-domain perception, real-time simulation, dynamic optimization and precise control of the production process. However, existing research mainly focuses on a single link or local application of concrete production, and there is still a lack of a full-process, multi-dimensional systematic solution. How to build a digital twin model covering the entire life cycle of concrete production, give full play to the comprehensive effectiveness of digital twins in production monitoring, scheduling optimization, quality control, etc., and realize the dynamic closed-loop optimization of manufacturing resource allocation and process execution still faces many technical challenges. Therefore, it is urgent to explore a new method of concrete production management based on digital twins to promote the upgrading of concrete production towards intelligence and leanness.

[0004] The purpose of the present invention is to provide a concrete production management method and system based on digital twins to address the problems of low efficiency and serious waste of resources in traditional concrete production management, so as to improve the intelligence level of concrete production management and achieve the goals of cost reduction, efficiency improvement and green manufacturing. Summary of the invention

[0005] The embodiments of the present invention provide a concrete production management method and system based on digital twin, which can solve the problems in the prior art.

[0006] According to a first aspect of the embodiments of the present invention,

[0007] A concrete production management method based on digital twin is provided, including:

[0008] The health monitoring data of concrete production equipment is collected in real time through a variety of sensors, and the full life cycle event data of concrete production equipment is collected using the IoT identification method. The collected health monitoring data and full life cycle event data are transmitted to the edge and cloud platforms respectively;

[0009] The health monitoring data is preprocessed at the edge using edge computing technology, and the state characteristics of the equipment are extracted. Based on the state characteristics of the equipment, a pre-trained fault prediction model is used to perform real-time diagnosis, and the annotation of the full life cycle event data is updated according to the diagnosis results. At the same time, the annotated full life cycle event data is used on the cloud platform to build a multi-level digital twin model of the equipment;

[0010] Based on the multi-level digital twin model, the pre-built physical degradation model and data-driven model are integrated to build an equipment remaining life prediction model. The equipment remaining life prediction model is used to predict the remaining life of the equipment. According to the prediction results, the concrete production strategy is dynamically adjusted in the digital twin model to obtain the optimal concrete production scheduling plan, which is converted into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan.

[0011] In an optional embodiment,

[0012] The edge computing technology is used at the edge to pre-process the health monitoring data and extract the status characteristics of the equipment. Based on the status characteristics of the equipment, a pre-trained fault prediction model is used to perform real-time diagnosis, including:

[0013] At the edge, the health monitoring data of concrete production equipment collected in real time by multiple sensors is received, and multi-domain time-frequency features that characterize the degradation state of the equipment are extracted. The attention mechanism is used to filter and obtain a subset of key features.

[0014] Based on the obtained key feature subsets, a multi-view learning strategy is introduced to construct multi-physical field feature subspaces, multi-sensor position feature subspaces, and multi-degradation feature subspaces for different degradation modes from three perspectives: physical field, sensor position, and degradation mode.

[0015] In each subspace, a multi-view attention mechanism is used to select the corresponding local key features, and different local key features are combined through an adaptive view weighting strategy to obtain a multi-view diagnostic feature set.

[0016] For the multi-view diagnostic feature set, a one-dimensional convolutional neural network is used at the edge to build an edge diagnostic model, which is used to determine whether the device is faulty.

[0017] If a fault occurs, a preliminary diagnosis of the fault mode is performed, and based on the preliminary diagnosis results, unknown complex faults with low confidence and the corresponding multi-view diagnostic features are uploaded to the cloud;

[0018] The cloud uses pre-trained fault diagnosis knowledge graphs to comprehensively consider equipment models, operating conditions, and degradation mechanisms to diagnose unknown complex faults. At the same time, the diagnosis results are used to optimize the fault diagnosis knowledge graph, and the optimized fault diagnosis knowledge graph is regularly embedded into the edge diagnosis model to guide edge diagnosis reasoning.

[0019] In an optional embodiment,

[0020] The cloud platform uses the labeled data of the entire life cycle of the equipment to build a multi-level digital twin model of the equipment, including:

[0021] Receiving the annotated equipment life cycle data in the cloud, obtaining the three-dimensional point cloud data therein, preprocessing the three-dimensional point cloud data, converting the preprocessed three-dimensional point cloud data into a triangular mesh surface model, and constructing a geometric model of the concrete production equipment using a surface fitting technology;

[0022] Based on the geometric model of concrete production equipment, the finite element analysis method is used to perform numerical simulation and solve the stress state, vibration characteristics and heat conduction characteristics of the concrete production equipment, obtain the physical response data of the concrete production equipment under different working conditions, and build the physical model of the equipment;

[0023] Obtain the condition monitoring parameters and performance parameters of concrete production equipment, use gated recurrent unit neural network to establish the nonlinear mapping relationship between condition monitoring parameters and performance parameters, and build an equipment degradation model for life prediction;

[0024] The equipment geometric model, equipment physical model and equipment degradation model are integrated to construct a multi-level digital twin model of concrete production equipment.

[0025] In an optional embodiment,

[0026] Based on the multi-level digital twin model, the pre-built equipment degradation model and data-driven model are integrated to build an equipment remaining life prediction model. The remaining life of the equipment is predicted using the equipment remaining life prediction model, including:

[0027] Based on the multi-level digital twin model, a data-driven model is constructed. The data-driven model uses a convolutional neural network as the basic structure and adaptively extracts multi-scale degradation features related to the equipment degradation process from health monitoring data through a multi-scale feature extraction mechanism.

[0028] Based on the obtained multi-scale degradation features, a joint loss function combining fault diagnosis loss and remaining life prediction loss is constructed, and the pre-constructed equipment degradation model is used as a regularization constraint term to train the data-driven model. The degradation trajectory output by the equipment degradation model is used as a priori constraint of the data-driven model to obtain prediction results that conform to the degradation trajectory, and the prediction results are used to feedback and optimize the equipment degradation model parameters. Repeat the iteration until the preset termination condition is met, and finally obtain a trained equipment remaining life prediction model.

[0029] Based on the trained equipment remaining life prediction model, the fault diagnosis results are used as prior knowledge, and the Bayesian inference method is used to predict the remaining life of the components by combining physical priors with data-driven posteriori.

[0030] In an optional embodiment,

[0031] Based on the obtained multi-scale degradation features, the calculation formula of the joint loss function combining fault diagnosis loss and remaining life prediction loss is as follows:

[0032]

[0033] Where L represents the joint loss function, α represents the weight coefficient of the fault diagnosis model loss, N represents the number of devices, M represents the number of fault categories, and y ij Indicates the true label of whether the i-th device belongs to the j-th type of fault, represents the predicted value of the mean of the characteristic vector of the i-th device under the j-th type of fault, μ j represents the true value of the mean of the characteristic vector of the jth type of fault, β represents the weight coefficient of the remaining life prediction model loss, W p (·) represents the difference between the remaining life distributions, represents the predicted remaining life distribution of the ith device, P ri represents the true remaining life distribution, γ represents the weight coefficient of the regularization term, W represents the parameter matrix of the model, and ||·||2 represents the L2 norm.

[0034] In an optional embodiment,

[0035] According to the prediction results, the concrete production strategy is dynamically adjusted in the digital twin model to obtain the optimal concrete production scheduling plan, which is converted into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan, including:

[0036] A multi-objective reinforcement learning optimization framework is constructed to model the concrete production process as a multi-objective Markov decision process, wherein the multi-objective Markov decision process includes a state space, an action space, a state transition probability matrix, a multi-objective reward function, a discount factor, and an initial state distribution, wherein the state space includes equipment state and process parameters, the action space includes equipment operation mode and production scheduling strategy, and the multi-objective reward function includes equipment remaining life, production efficiency, and energy consumption indicators;

[0037] Based on the constructed multi-objective Markov decision process, a hierarchical target learning algorithm is designed, and a value function approximator and a policy function approximator are constructed for each optimization target of the multi-objective reward function. The value function and the policy function of each optimization target are estimated by using the value function approximator and the policy function approximator. According to the obtained value function estimation value and policy function estimation value, a target improvement operator is obtained, and the multi-objective reward function is updated based on the target improvement operator. Based on the updated multi-objective reward function, a non-dominated sorting genetic algorithm is used to search in the multi-objective Markov decision process to obtain an initial policy set.

[0038] Based on the initial strategy set obtained, in the digital twin model, by building an interactive interface between the intelligent agent and the virtual environment, the Monte Carlo tree search algorithm is used to select the optimal exploration path through an adaptive exploration-utilization equilibrium strategy, generate large-scale optimization strategy sample data, and obtain the optimal strategy set;

[0039] The optimal strategy set is input into the interactive decision recommendation system of the target learning algorithm, and the feasibility of the strategy is evaluated in combination with expert knowledge to generate the final optimal concrete production scheduling plan;

[0040] The optimal concrete production scheduling plan is converted into equipment control instructions, and the automatic issuance and execution of the concrete production scheduling plan is realized through the interactive interface between the digital twin model and the actual production management system.

[0041] In an optional embodiment,

[0042] A value function approximator and a policy function approximator are constructed for each optimization objective of the multi-objective reward function. The value function and the policy function of each optimization objective are estimated using the value function approximator and the policy function approximator. The target improvement operator is obtained according to the obtained value function estimation value and policy function estimation value, and the calculation formula for updating the multi-objective reward function based on the target improvement operator is as follows:

[0043]

[0044] Among them, R′ k represents the updated multi-objective reward function, R krepresents the initial multi-objective reward function, λ represents the learning rate, A represents the action space, δ represents the discount factor, S represents the state space, P(·) represents the state transition probability matrix, represents the value function of the kth optimization objective estimated by the value function approximator at state s′, represents the value function of the kth optimization objective estimated by the value function approximator at state s, ω represents the adjustment factor that balances the reward function and the policy function, represents the probability of the kth optimization objective choosing action a′ in state s estimated by the policy function approximator, represents the probability of selecting action a in state s as estimated by the policy function approximator for the kth optimization objective.

[0045] According to a second aspect of the embodiments of the present invention,

[0046] Provide a concrete production management system based on digital twin, including:

[0047] The first unit is used to collect health monitoring data of concrete production equipment in real time through multiple sensors, and collect full life cycle event data of concrete production equipment using IoT identification methods, and transmit the collected health monitoring data and full life cycle event data to the edge and cloud platforms respectively;

[0048] The second unit is used to pre-process the health monitoring data using edge computing technology at the edge end, and extract the status characteristics of the equipment, perform real-time diagnosis based on the status characteristics of the equipment using a pre-trained fault prediction model, and update the annotation of the full life cycle event data according to the diagnosis results, and at the same time use the annotated full life cycle event data on the cloud platform to build a multi-level digital twin model of the equipment;

[0049] The third unit is used to integrate the pre-built physical degradation model and data-driven model based on the multi-level digital twin model, build an equipment remaining life prediction model, use the equipment remaining life prediction model to predict the remaining life of the equipment, and dynamically adjust the concrete production strategy in the digital twin model according to the prediction results to obtain the optimal concrete production scheduling plan, and convert it into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan.

[0050] According to a third aspect of the embodiments of the present invention,

[0051] An electronic device is provided, comprising:

[0052] processor;

[0053] a memory for storing processor-executable instructions;

[0054] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0055] A fourth aspect of the embodiments of the present invention is:

[0056] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0057] In this embodiment, the health monitoring data of the equipment is collected in real time through a variety of sensors, and the whole life cycle event data is collected by using the IoT identification method, which can fully perceive and record the operating status, fault events, maintenance records, etc. of the equipment, providing complete data support for equipment health management. Using edge computing technology, real-time data preprocessing and fault diagnosis can be performed on the edge side close to the equipment, which improves the response speed and real-time performance and reduces the computing pressure on the cloud. Through the Bayesian inference framework, the pre-built physical degradation model is integrated with the data-driven model based on monitoring data to form a hybrid model for remaining life prediction, which makes full use of prior knowledge and data posteriori, and improves the accuracy and reliability of the prediction. Based on the remaining life prediction results of the equipment, combined with advanced technologies such as reinforcement learning and evolutionary algorithms, the concrete production scheduling strategy is dynamically optimized in the digital twin model, and the optimal production scheduling plan is obtained under the premise of meeting multiple objective constraints, which improves production efficiency, saves energy and extends equipment life, and provides intelligent, automated and optimized solutions for complex industrial processes such as concrete production. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a flow chart of a concrete production management method based on digital twin according to an embodiment of the present invention;

[0059] Figure 2 It is a structural schematic diagram of a concrete production management system based on digital twin according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0062] Figure 1 Schematic diagram of the process of the concrete production management method based on digital twin according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0063] S 1 01. Collect health monitoring data of concrete production equipment in real time through multiple sensors, and use the Internet of Things identification method to collect full life cycle event data of concrete production equipment, and transmit the collected health monitoring data and full life cycle event data to the edge and cloud platforms respectively.

[0064] For example, multi-sensor fusion and industrial Internet of Things technology are first used to collect multi-dimensional and full-life cycle data on the operating status of the equipment. Vibration, temperature, pressure, strain and other sensors are deployed in key parts of the equipment, such as bearings, reducers, motors, etc., to collect dynamic response data of the equipment in real time. Using Internet of Things identification technologies such as radio frequency identification (RFID) and data matrix codes, data is recorded and associated with full life cycle events such as manufacturing, assembly, and maintenance of equipment to achieve interconnection and interoperability of data throughout the equipment life cycle. At the same time, through communication technologies such as industrial Ethernet and wireless sensor networks, multi-source heterogeneous equipment status data is connected to the cloud platform to build a big data center for the entire life cycle of the equipment.

[0065] In this embodiment, equipment health monitoring data is collected in real time through a variety of sensors, and full life cycle event data is acquired using IoT identification technology. This allows for comprehensive perception of the operating status, failure mode, maintenance history, etc. of concrete production equipment, providing complete data support for equipment health management.

[0066] S102. Preprocess the health monitoring data using edge computing technology at the edge, extract the status characteristics of the equipment, perform real-time diagnosis based on the status characteristics of the equipment using a pre-trained fault prediction model, and update the annotation of the full life cycle event data based on the diagnosis results. At the same time, use the annotated full life cycle event data on the cloud platform to build a multi-level digital twin model of the equipment.

[0067] Exemplarily, edge computing technology is used to realize real-time data processing and feature extraction at the data collection end. Edge computing devices such as smart gateways and embedded systems are deployed at the production site to pre-process the collected raw sensor data such as vibration and temperature, and extract health monitoring features such as time domain, frequency domain, and time-frequency domain. Signal processing methods such as wavelet transform and empirical mode decomposition are used to denoise and enhance vibration signals and extract features. Statistical indicators such as kurtosis and root mean square are used to characterize the degradation trend of the health status of the equipment. Using end-side reasoning technology based on deep learning, pre-trained convolutional neural network models are deployed on gateway devices to realize real-time diagnosis and early warning of equipment failure modes.

[0068] When the fault diagnosis model on the edge outputs the result, a corresponding confidence score will be given to represent the credibility of the fault type judgment. A confidence threshold is set. If the confidence of the diagnosis result is higher than the threshold, it is considered a "reliable result" and can be directly used to update the event data annotation; if the confidence is lower than the threshold, it is considered an "uncertain result" and needs to be uploaded to the cloud for further diagnosis and manual confirmation.

[0069] For "reliable results", the edge will associate the time point, fault type, location and other information of the fault with the device ID, construct a label string, upload the constructed label string to the cloud, and update the device's full life cycle event data file stored in the cloud; at the same time, for "uncertain results", the edge will upload the corresponding feature data, preliminary diagnosis results, etc. to the cloud. The cloud uses more complex models for in-depth diagnosis and combines the experience of manual experts to make judgments to obtain confirmed fault types.

[0070] The confirmed fault information is also constructed as a label string and updated to the full life cycle event data file. It can continuously record various events and faults occurring in the equipment in the full life cycle event data, laying a solid data foundation for building a high-fidelity digital twin model.

[0071] In an optional embodiment,

[0072] The edge computing technology is used at the edge to pre-process the health monitoring data and extract the status characteristics of the equipment. Based on the status characteristics of the equipment, a pre-trained fault prediction model is used to perform real-time diagnosis, including:

[0073] At the edge, the health monitoring data of concrete production equipment collected in real time by multiple sensors is received, and multi-domain time-frequency features that characterize the degradation state of the equipment are extracted. The attention mechanism is used to filter and obtain a subset of key features.

[0074] Based on the obtained key feature subsets, a multi-view learning strategy is introduced to construct multi-physical field feature subspaces, multi-sensor position feature subspaces, and multi-degradation feature subspaces for different degradation modes from three perspectives: physical field, sensor position, and degradation mode.

[0075] In each subspace, a multi-view attention mechanism is used to select the corresponding local key features, and different local key features are combined through an adaptive view weighting strategy to obtain a multi-view diagnostic feature set.

[0076] For the multi-view diagnostic feature set, a one-dimensional convolutional neural network is used at the edge to build an edge diagnostic model, which is used to determine whether the device is faulty.

[0077] If a fault occurs, a preliminary diagnosis of the fault mode is performed, and based on the preliminary diagnosis results, unknown complex faults with low confidence and the corresponding multi-view diagnostic features are uploaded to the cloud;

[0078] The cloud uses pre-trained fault diagnosis knowledge graphs to comprehensively consider equipment models, operating conditions, and degradation mechanisms to diagnose unknown complex faults. At the same time, the diagnosis results are used to optimize the fault diagnosis knowledge graph, and the optimized fault diagnosis knowledge graph is regularly embedded into the edge diagnosis model to guide edge diagnosis reasoning.

[0079] Exemplarily, health monitoring data is first collected in real time from various sensors installed in concrete production equipment (such as vibration sensors, temperature sensors, pressure sensors, etc.), including time series data, spectrum data, etc. The raw data is preprocessed, such as denoising, interpolation, etc., to eliminate interference and invalid data.

[0080] Extract multi-domain time-frequency features that characterize the degradation state of the equipment, such as time-domain features, such as mean, peak, peak-to-peak, square root amplitude and other statistics; frequency-domain features, such as power spectrum density, frequency center, band power, etc.; time-frequency domain features, such as wavelet packet decomposition energy, wavelet energy entropy, etc. Use the attention mechanism to perform weighted fusion on the extracted multi-domain features to obtain a subset of key features.

[0081] Based on the key feature subsets, a multi-view learning strategy is introduced. From the perspective of physical fields, a multi-physical field feature subspace is constructed, and the features corresponding to different physical fields (such as vibration, temperature, pressure, etc.) are grouped, modeled and learned separately, and unsupervised methods such as autoencoders are used to learn feature representation for each physical field subspace. From the perspective of sensor position, a multi-sensor position feature subspace is constructed, and the features corresponding to different sensor positions are grouped. Feature learning is performed on each position subspace to capture position-related degradation patterns. From the perspective of degradation patterns, multiple degradation feature subspaces are constructed, and features are grouped according to known degradation patterns (such as wear, cracks, corrosion, etc.), and a special feature extraction model is trained for each degradation pattern subspace.

[0082] In each feature subspace, the multi-view attention mechanism is used to screen the corresponding local key features, and the local key features of different perspectives are combined through the adaptive view weighting strategy to obtain the multi-view diagnostic feature set. Specifically, for each feature subspace, the self-attention mechanism is used to capture the long-range dependencies between features. The attention weights of each feature and other features are calculated, and the high-weight features are used as local key features. The gated recurrent unit (GRU) is used to perform temporal modeling on the local key features of each perspective. Adaptive weights are assigned to different perspectives through the view attention mechanism. The weighted local key features are spliced ​​to obtain the multi-view diagnostic feature set.

[0083] At the edge, a one-dimensional convolutional neural network is used to build an edge diagnosis model. The input layer receives the multi-view diagnosis feature set, the convolution layer includes multiple layers of 1D convolutional layers, each layer contains convolution, batch normalization and activation functions (such as ReLU), the pooling layer is used for feature downsampling and translation invariance, the fully connected layer performs nonlinear transformation and mapping on the features, and the output layer sets the number of nodes according to the number of fault types, such as 1 node for binary classification and multiple nodes for multi-classification. The multi-view diagnosis feature set is used as the input of the one-dimensional convolutional neural network, and after multiple layers of convolution and pooling operations, the fault diagnosis result is output. The fault diagnosis model is trained by supervised learning, and the model is optimized using labeled data. The trained fault diagnosis model is used to perform online fault diagnosis on the real-time status of the equipment.

[0084] If the edge diagnosis model determines that a fault has occurred, but the confidence level is low (i.e., it is an unknown complex fault), the multi-view diagnostic features are uploaded to the cloud. The cloud uses the pre-trained fault diagnosis knowledge graph, combined with information such as equipment model, operating condition, and degradation mechanism, to diagnose unknown complex faults. The diagnostic knowledge graph contains knowledge such as equipment failure mode, failure characteristics, and diagnostic logic, where node types include equipment model, component, operating condition, degradation mode, failure mode, etc. The edge relationship types include "contains", "leads to", "manifests as", etc. According to the multi-view features uploaded by the edge, the best matching fault mode path is found, and the knowledge graph reasoning module (such as GNNs) is used to infer the cause of the fault and the propagation path.

[0085] After the reasoning is completed, new fault cases are added to the knowledge graph to continuously enrich the knowledge base. According to the diagnosis effect, the relationship weights in the knowledge graph are adjusted. At the same time, a federated transfer learning strategy is adopted to use the diagnostic knowledge from different devices and different scenarios in the cloud to continuously expand and optimize the knowledge graph, and regularly embed the graph into the edge diagnosis model to guide edge diagnosis reasoning.

[0086] Optionally, for samples with low diagnostic confidence at the edge, the most valuable samples are adaptively labeled by maximizing diagnostic entropy, and incremental labeled samples are regularly uploaded to the cloud for fine-tuning of the diagnostic model. Under the federated learning framework, different edge nodes use mechanisms such as encrypted communication and differential privacy to achieve secure sharing of incremental samples and model parameters. Without exposing the original data, the labeled data of other nodes are used to enhance the discrimination ability of the local diagnostic model. The cloud uses the federated distillation algorithm to compress the globally aggregated incremental samples into a small-scale distillation set, selectively inherit the optimization results of distributed incremental training, improve the generalization of the cloud diagnostic model while ensuring privacy and security, and transmit the optimized diagnostic knowledge back to the edge to achieve the co-evolution of the end-cloud diagnostic model. The incremental training framework introduces active learning at the edge to achieve autonomous expansion of diagnostic samples; it integrates federated learning in the cloud to achieve distributed sharing of knowledge, and improves the efficiency and generalization of distributed diagnostic model learning in complex industrial scenarios.

[0087] In this embodiment, edge computing technology is used to process monitoring data and diagnose faults at the nearest edge node, which greatly improves the real-time and response speed of diagnosis. Extracting key features based on attention mechanism and multi-view learning strategy is conducive to accurately capturing the degradation state of the equipment and improving the accuracy of fault diagnosis. Most common faults are diagnosed at the edge, and key features are uploaded to the cloud only when unknown complex faults are encountered, which greatly reduces the amount of data transmission. The pre-trained fault diagnosis knowledge graph is used in the cloud, which integrates knowledge in the fields of equipment model, working condition, degradation mechanism, etc., improves the diagnostic ability of unknown complex faults, and can give full play to the different advantages of the edge and cloud to realize intelligent and efficient fault diagnosis.

[0088] In an optional embodiment,

[0089] The cloud platform uses the labeled data of the entire life cycle of the equipment to build a multi-level digital twin model of the equipment, including:

[0090] Receiving the annotated equipment life cycle data in the cloud, obtaining the three-dimensional point cloud data therein, preprocessing the three-dimensional point cloud data, converting the preprocessed three-dimensional point cloud data into a triangular mesh surface model, and constructing a geometric model of the concrete production equipment using a surface fitting technology;

[0091] Based on the geometric model of concrete production equipment, the finite element analysis method is used to perform numerical simulation and solve the stress state, vibration characteristics and heat conduction characteristics of the concrete production equipment, obtain the physical response data of the concrete production equipment under different working conditions, and build the physical model of the equipment;

[0092] Obtain the condition monitoring parameters and performance parameters of concrete production equipment, use gated recurrent unit neural network to establish the nonlinear mapping relationship between condition monitoring parameters and performance parameters, and build an equipment degradation model for life prediction;

[0093] The equipment geometric model, equipment physical model and equipment degradation model are integrated to construct a multi-level digital twin model of concrete production equipment.

[0094] Exemplarily, first obtain three-dimensional point cloud data from the equipment's full life cycle data, and preprocess the original point cloud data, including denoising, filling, downsampling, etc., to eliminate outliers and redundant data. Use geometric processing algorithms, such as spherical projection, RANSAC, etc., to segment the preprocessed point cloud data and extract the main components of the equipment. For each component, use surface reconstruction algorithms, such as moving least squares, Gaussian process regression, etc., to fit the triangular mesh surface model. Assemble the surface models of each component to construct a three-dimensional geometric model of the entire concrete production equipment.

[0095] Based on the geometric model, the finite element analysis software is used to model the physical characteristics of the equipment. Define the material properties and boundary conditions of the equipment. The material properties of the equipment include density, elastic modulus, thermal conductivity, etc. Perform static analysis on the stress state of the equipment, and calculate the stress and strain distribution. Perform modal analysis on the vibration characteristics of the equipment to solve the natural frequency, modal shape, etc. Perform thermal analysis on the heat conduction characteristics of the equipment, calculate the temperature field distribution, and then export the physical response data (such as stress, displacement, temperature, etc.) under different working conditions to build a physical model of the equipment.

[0096] Collect the equipment's condition monitoring parameters (such as vibration, temperature, pressure, etc.) and performance parameters (such as production capacity, energy consumption, failure rate, etc.). Use the gated recurrent unit neural network to establish the mapping relationship between the condition monitoring parameters and the performance parameters. Among them, the GRU network structure includes an input layer, multi-layer GRU units, a fully connected layer, and an output layer. Use the condition monitoring parameter sequence as the GRU input and the performance parameters as the supervision label to train the model. The training goal is to minimize the prediction error of the performance parameters and obtain the optimal mapping model. Use the trained GRU network model to predict the performance parameters of the equipment based on the condition monitoring parameters and build a degradation model.

[0097] The equipment geometry model, physical model and degradation model are integrated into a unified model framework, and the data interaction between the three sub-models is realized by using collaborative simulation technology. Among them, the geometry model provides 3D visualization and spatial information of the equipment, and the physical model calculates the physical response of the equipment according to the working conditions as the input of the degradation model. The degradation model predicts the performance parameters of the equipment based on the physical response and feeds back to the physical model to update the working conditions. The three sub-models interact and iterate to form a digital twin of the equipment throughout its life cycle.

[0098] In this embodiment, the accurate geometric shape of the equipment is reconstructed using three-dimensional point cloud data, and a fine geometric digital twin model is constructed through surface fitting technology, which lays the foundation for subsequent physical simulation and visualization analysis. Based on the finite element analysis method, the physical processes such as the stress state, vibration characteristics and heat conduction of the equipment are accurately simulated, which can truly reproduce the physical response of the equipment under different working conditions and provide support for safety and reliability analysis. Using deep learning technologies such as gated recurrent unit neural networks, a nonlinear mapping relationship between equipment status and performance parameters is established from monitoring data, which can predict the aging degree and remaining life of the equipment and provide a basis for formulating preventive maintenance plans. The geometric model, physical model and degradation model are organically integrated to construct a digital twin model of concrete production equipment that integrates multi-level knowledge. It not only has visualization and physical simulation functions, but also can predict the life of the equipment and support the full life cycle management of the equipment.

[0099] S 1 03. Based on the multi-level digital twin model, the pre-built equipment degradation model and data-driven model are integrated to build an equipment remaining life prediction model. The equipment remaining life prediction model is used to predict the remaining life of the equipment. According to the prediction results, the concrete production strategy is dynamically adjusted in the digital twin model to obtain the optimal concrete production scheduling plan, which is converted into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan.

[0100] In an optional embodiment,

[0101] Based on the multi-level digital twin model, the pre-built equipment degradation model and data-driven model are integrated to build an equipment remaining life prediction model. The remaining life of the equipment is predicted using the equipment remaining life prediction model, including:

[0102] Based on the multi-level digital twin model, a data-driven model is constructed. The data-driven model uses a convolutional neural network as the basic structure and adaptively extracts multi-scale degradation features related to the equipment degradation process from health monitoring data through a multi-scale feature extraction mechanism.

[0103] Based on the obtained multi-scale degradation features, a joint loss function combining fault diagnosis loss and remaining life prediction loss is constructed, and the pre-constructed equipment degradation model is used as a regularization constraint term to train the data-driven model. The degradation trajectory output by the equipment degradation model is used as a priori constraint of the data-driven model to obtain prediction results that conform to the degradation trajectory, and the prediction results are used to feedback and optimize the equipment degradation model parameters. Repeat the iteration until the preset termination condition is met, and finally obtain a trained equipment remaining life prediction model.

[0104] Based on the trained equipment remaining life prediction model, the fault diagnosis results are used as prior knowledge, and the Bayesian inference method is used to predict the remaining life of the components by combining physical priors with data-driven posteriori.

[0105] Exemplarily, a data-driven model is first constructed based on the digital twin model. The data-driven model has a convolutional neural network as the basic structure, which includes multiple convolutional layers, pooling layers, and fully connected layers, and is used to automatically extract features from the original input. A multi-scale feature extraction mechanism is introduced to capture degradation features of different scales through convolution kernels of different sizes, and the health monitoring data of the equipment (such as vibration, temperature, etc.) is used as the input of the data-driven model. Through end-to-end training, multi-scale features related to the physical degradation process are adaptively learned.

[0106] A joint loss function is designed, which includes two parts: fault diagnosis loss and remaining life prediction loss. The fault diagnosis loss is based on classification cross entropy loss, which is used to distinguish between faults and normal states. The remaining life prediction loss uses regression loss functions such as mean square error or smooth L1 loss. The output of the pre-built equipment degradation model is introduced as a regularization constraint term into the joint loss function. By minimizing the joint loss function, the training data-driven model can simultaneously meet the goals of fault diagnosis and life prediction.

[0107] The equipment degradation model is iteratively integrated with the data-driven model. Specifically, the parameters of the equipment degradation model are initialized, pre-trained based on existing knowledge or historical data, and the degradation trajectory output by the equipment degradation model is used as the prior constraint of the data-driven model. When training the data-driven model, the weight of the loss function is adjusted according to the degradation trajectory to optimize the model output, and the parameters of the equipment degradation model are updated using the prediction results of the data-driven model. The above iterative process is repeated until the model converges or meets the preset termination conditions, and finally a hybrid degradation model that integrates the physical model and the data-driven model is obtained.

[0108] The fault type, occurrence time and other information of the equipment are obtained from the fault diagnosis module, and the fault information is encoded into a probability distribution as one of the prior distributions of the Bayesian inference framework. For example, the fault type can be encoded with one-hot, and the fault occurrence time can be modeled with a Gaussian distribution. The equipment degradation model describes the degradation process of the equipment under different working conditions and material properties. According to the equipment model, process parameters, etc., the parameter value or distribution of the degradation model is determined, and these physical parameters and their distribution are used as another prior distribution of the Bayesian inference framework.

[0109] The trained data-driven model is used to predict the remaining life of new monitoring data. The output of the data-driven model is a point estimate of the remaining life or a parameterized distribution (such as a Gaussian distribution). The prediction results of the data-driven model are used as the likelihood function or posterior distribution of the Bayesian inference framework.

[0110] According to Bayes' theorem, the fault prior, physical prior and data posterior are combined. Since the posterior distribution often has no analytical solution, a sampling method is needed to approximate the solution. Markov Monte Carlo sampling (MCMC) is a commonly used sampling method. Through MCMC, a large number of remaining life samples are extracted from the posterior distribution to approximately reconstruct the posterior distribution.

[0111] From the posterior distribution obtained by sampling, the mean or median of the remaining life is calculated as a point estimate. According to the distribution of the sample, the confidence interval of the remaining life, such as a 95% credible interval, is determined. The confidence interval can quantify the uncertainty of the prediction and provide a basis for subsequent decision-making.

[0112] In an optional embodiment,

[0113] Based on the obtained multi-scale degradation features, the calculation formula of the joint loss function combining fault diagnosis loss and remaining life prediction loss is as follows:

[0114]

[0115] Where L represents the joint loss function, α represents the weight coefficient of the fault diagnosis model loss, N represents the number of devices, M represents the number of fault categories, and y ij Indicates the true label of whether the i-th device belongs to the j-th type of fault, represents the predicted value of the feature vector mean of the i-th device under the j-th type of fault, μ j represents the true value of the mean of the characteristic vector of the jth type of fault, β represents the weight coefficient of the remaining life prediction model loss, W p (·) represents the difference between the remaining life distributions, represents the predicted remaining life distribution of the ith device, P ri represents the true remaining life distribution, γ represents the weight coefficient of the regularization term, W represents the parameter matrix of the model, and ||·||2 represents the L2 norm.

[0116] In this embodiment, the degradation model based on physical process and the driving model based on data are integrated, and the advantages of both are utilized to more accurately describe and predict the degradation process of the equipment. The data-driven model adaptively extracts multi-scale degradation features from the monitoring data through convolutional neural networks, which helps to mine complex nonlinear degradation patterns. The degradation model provides physical constraints to ensure that the prediction results conform to the degradation trajectory and avoid the deviation and outliers of the pure data-driven model. Using the Bayesian inference framework, the pre-built physical degradation model is used as prior knowledge, and the data-driven model output is used as a posteriori evidence. By effectively integrating physical priors with data posteriori, the stability and reliability of the prediction can be effectively improved. By constructing a joint loss function, the fault diagnosis loss and the remaining life prediction loss are organically combined, and the resulting hybrid degradation model can describe complex nonlinear degradation processes and improve the generalization performance of the prediction. This technical solution provides an integrated life prediction solution, provides technical support for the full life cycle management of equipment, thereby promoting the intelligent upgrading of the manufacturing industry, improving equipment utilization efficiency and reliability, and reducing operation and maintenance costs.

[0117] In an optional embodiment,

[0118] According to the prediction results, the concrete production strategy is dynamically adjusted in the digital twin model to obtain the optimal concrete production scheduling plan, which is converted into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan, including:

[0119] A multi-objective reinforcement learning optimization framework is constructed to model the concrete production process as a multi-objective Markov decision process, wherein the multi-objective Markov decision process includes a state space, an action space, a state transition probability matrix, a multi-objective reward function, a discount factor, and an initial state distribution, wherein the state space includes equipment state and process parameters, the action space includes equipment operation mode and production scheduling strategy, and the multi-objective reward function includes equipment remaining life, production efficiency, and energy consumption indicators;

[0120] Based on the constructed multi-objective Markov decision process, a hierarchical target learning algorithm is designed, and a value function approximator and a policy function approximator are constructed for each optimization target of the multi-objective reward function. The value function and the policy function of each optimization target are estimated by using the value function approximator and the policy function approximator. According to the obtained value function estimation value and policy function estimation value, a target improvement operator is obtained, and the multi-objective reward function is updated based on the target improvement operator. Based on the updated multi-objective reward function, a non-dominated sorting genetic algorithm is used to search in the multi-objective Markov decision process to obtain an initial policy set.

[0121] Based on the initial strategy set obtained, in the digital twin model, by building an interactive interface between the intelligent agent and the virtual environment, the Monte Carlo tree search algorithm is used to select the optimal exploration path through an adaptive exploration-utilization equilibrium strategy, generate large-scale optimization strategy sample data, and obtain the optimal strategy set;

[0122] The optimal strategy set is input into the interactive decision recommendation system of the target learning algorithm, and the feasibility of the strategy is evaluated in combination with expert knowledge to generate the final optimal concrete production scheduling plan;

[0123] The optimal concrete production scheduling plan is converted into equipment control instructions, and the automatic issuance and execution of the concrete production scheduling plan is realized through the interactive interface between the digital twin model and the actual production management system.

[0124] Exemplarily, a multi-objective reinforcement learning optimization framework is first constructed to model the concrete production process as a multi-objective Markov decision process. Define the state space, including equipment status (such as temperature, vibration, etc.) and process parameters (such as material properties, environmental conditions, etc.). Define the action space, including equipment operation modes (such as on / off, load adjustment, etc.) and production scheduling strategies (such as capacity adjustment, process optimization, etc.). Determine the state transition probability matrix, which describes the probability distribution of transitioning from the current state to the next state after taking a certain action, which can be estimated based on a physical model or historical data. Construct a multi-objective reward function, including indicators such as remaining life of equipment, production efficiency, and energy consumption, and set weights according to priority. Set a discount factor to balance immediate rewards and long-term cumulative rewards. Determine the initial state distribution to describe the starting state of the production task.

[0125] Design a hierarchical multi-objective learning algorithm, and build corresponding value function approximators and policy function approximators for each optimization objective. The value function approximator can use a deep neural network structure, such as a fully connected network or a long short-term memory network. The policy function approximator can also use a deep neural network, and the output is an action probability distribution or a deterministic action.

[0126] Based on temporal difference or policy gradient estimation, the value function and policy function are specifically estimated. For each target, the temporal difference learning algorithm or policy gradient algorithm is used to update the parameters of the value function approximator and the policy function approximator through the state-action-reward sequence data generated by interaction with the virtual environment. Techniques such as experience replay and target network can also be used to improve training stability and convergence.

[0127] According to the estimated value function and policy function, the target improvement operator (such as gradient or Monte Carlo estimation) corresponding to each target is calculated. The improvement operators of each target are linearly combined according to the weights to form a multi-target improvement operator. The multi-target improvement operator is used to update the multi-target reward function through policy gradient ascent or other optimization methods.

[0128] Multi-objective optimization is performed based on the updated reward function, and optionally, a non-dominated sorting genetic algorithm is used to optimize the updated multi-objective reward function. The non-dominated sorting genetic algorithm performs selection, crossover and mutation operations according to the non-dominated sorting and crowding distance of individuals in the population, and obtains an initial strategy set that can achieve a balance between multiple objectives through iterative evolution.

[0129] The digital twin model is used as a virtual environment in which the agent can observe the state and perform actions. The state observation interface and action execution interface are defined to achieve two-way data exchange. The agent selects an action based on the current state, and the environment updates the next state based on the action and returns the reward. The Monte Carlo tree search algorithm is embedded in the decision-making mechanism of the agent. In the Monte Carlo tree, each node represents a state, and the edge represents the action and its transition probability. An adaptive exploration-utilization equilibrium strategy (such as UCB or progressive knowledge utilization) is used to select the optimal path to expand the tree node. Through a large number of simulated interactions, the optimization strategy sample data is generated, and the strategy set with the best performance is selected.

[0130] The optimized strategy set is input into the interactive decision recommendation system of the target learning algorithm, and the feasibility, risk and constraints of each strategy are evaluated in combination with the domain knowledge of artificial experts. Based on the evaluation results, the final optimal concrete production scheduling solution is selected from the strategy set.

[0131] The optimal scheduling plan is converted into a specific sequence of equipment control instructions. Through the two-way interactive interface between the digital twin model and the actual production management system, the control instructions are automatically sent to the execution system at the production site, realizing real-time execution of the scheduling plan, real-time monitoring of the production process, and dynamic adjustment of the scheduling strategy based on feedback information.

[0132] In an optional embodiment,

[0133] A value function approximator and a policy function approximator are constructed for each optimization objective of the multi-objective reward function. The value function and the policy function of each optimization objective are estimated using the value function approximator and the policy function approximator. The target improvement operator is obtained according to the obtained value function estimation value and policy function estimation value, and the calculation formula for updating the multi-objective reward function based on the target improvement operator is as follows:

[0134]

[0135] Among them, R′ k represents the updated multi-objective reward function, R k represents the initial multi-objective reward function, λ represents the learning rate, A represents the action space, δ represents the discount factor, S represents the state space, P(·) represents the state transition probability matrix, represents the value function of the kth optimization objective estimated by the value function approximator at state s′, represents the value function of the kth optimization objective estimated by the value function approximator at state s, ω represents the adjustment factor that balances the reward function and the policy function, represents the probability of the kth optimization objective choosing action a′ in state s estimated by the policy function approximator, represents the probability of selecting action a in state s for the kth optimization objective estimated by the policy function approximator.

[0136] In this embodiment, the concrete production process is modeled as a multi-objective Markov decision process, which comprehensively considers multiple optimization objectives such as the remaining life of the equipment, production efficiency and energy consumption. It can optimize multiple objective reward functions at the same time and obtain the optimal decision that balances multiple objectives. Advanced algorithms such as Monte Carlo tree search and non-dominated sorting genetic algorithm are integrated to generate large-scale optimization samples through interactive exploration between intelligent agents and virtual environments, and can generate highly intelligent optimal decision plans. The state of the Markov decision process is dynamically updated according to the latest information, and the scheduling plan is optimized in real time to achieve dynamic adaptive optimization of production scheduling. The dynamically optimized scheduling decision plan can be directly converted into equipment control instructions to achieve automated control of the production process. Through modeling, simulation, optimization, and control instruction issuance, the highly intelligent digital twin model is integrated into the entire actual concrete production process, demonstrating the huge application potential of digital twin technology in intelligent manufacturing, which will promote the intelligent upgrading of the manufacturing industry.

[0137] Figure 2 Schematic diagram of the structure of a concrete production management system based on digital twins according to an embodiment of the present invention. Figure 2 As shown, the system comprises:

[0138] The first unit is used to collect health monitoring data of concrete production equipment in real time through multiple sensors, and collect full life cycle event data of concrete production equipment using IoT identification methods, and transmit the collected health monitoring data and full life cycle event data to the edge and cloud platforms respectively;

[0139] The second unit is used to pre-process the health monitoring data using edge computing technology at the edge end, and extract the status characteristics of the equipment, perform real-time diagnosis based on the status characteristics of the equipment using a pre-trained fault prediction model, and update the annotation of the full life cycle event data according to the diagnosis results, and at the same time use the annotated full life cycle event data on the cloud platform to build a multi-level digital twin model of the equipment;

[0140] The third unit is used to integrate the pre-built physical degradation model and data-driven model based on the multi-level digital twin model, build an equipment remaining life prediction model, use the equipment remaining life prediction model to predict the remaining life of the equipment, and dynamically adjust the concrete production strategy in the digital twin model according to the prediction results to obtain the optimal concrete production scheduling plan, and convert it into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan.

[0141] According to a third aspect of the embodiments of the present invention,

[0142] An electronic device is provided, comprising:

[0143] processor;

[0144] a memory for storing processor-executable instructions;

[0145] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0146] A fourth aspect of the embodiments of the present invention is:

[0147] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0148] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A concrete production management method based on digital twin, characterized in that: include: The health monitoring data of concrete production equipment is collected in real time through a variety of sensors, and the full life cycle event data of concrete production equipment is collected using the IoT identification method. The collected health monitoring data and full life cycle event data are transmitted to the edge and cloud platforms respectively; The health monitoring data is preprocessed at the edge using edge computing technology, and the state characteristics of the equipment are extracted. Based on the state characteristics of the equipment, a pre-trained fault prediction model is used to perform real-time diagnosis, and the annotation of the full life cycle event data is updated according to the diagnosis results. At the same time, the annotated full life cycle event data is used on the cloud platform to build a multi-level digital twin model of the equipment; Based on the multi-level digital twin model, the pre-built physical degradation model and data-driven model are integrated to build an equipment remaining life prediction model. The equipment remaining life prediction model is used to predict the remaining life of the equipment. According to the prediction results, the concrete production strategy is dynamically adjusted in the digital twin model to obtain the optimal concrete production scheduling plan, which is converted into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan.

2. The method according to claim 1, characterized in that The edge computing technology is used at the edge to pre-process the health monitoring data and extract the status characteristics of the equipment. Based on the status characteristics of the equipment, a pre-trained fault prediction model is used to perform real-time diagnosis, including: At the edge, the health monitoring data of concrete production equipment collected in real time by multiple sensors is received, and multi-domain time-frequency features that characterize the degradation state of the equipment are extracted. The attention mechanism is used to filter and obtain a subset of key features. Based on the obtained key feature subsets, a multi-view learning strategy is introduced to construct multi-physical field feature subspaces, multi-sensor position feature subspaces, and multi-degradation feature subspaces for different degradation modes from three perspectives: physical field, sensor position, and degradation mode. In each subspace, a multi-view attention mechanism is used to select the corresponding local key features, and different local key features are combined through an adaptive view weighting strategy to obtain a multi-view diagnostic feature set. For the multi-view diagnostic feature set, a one-dimensional convolutional neural network is used at the edge to build an edge diagnostic model, which is used to determine whether the device is faulty. If a fault occurs, a preliminary diagnosis of the fault mode is performed, and based on the preliminary diagnosis results, unknown complex faults with low confidence and the corresponding multi-view diagnostic features are uploaded to the cloud; The cloud uses pre-trained fault diagnosis knowledge graphs to comprehensively consider equipment models, operating conditions, and degradation mechanisms to diagnose unknown complex faults. At the same time, the diagnosis results are used to optimize the fault diagnosis knowledge graph, and the optimized fault diagnosis knowledge graph is regularly embedded into the edge diagnosis model to guide edge diagnosis reasoning.

3. The method according to claim 1, characterized in that The cloud platform uses the labeled data of the entire life cycle of the equipment to build a multi-level digital twin model of the equipment, including: Receiving the annotated equipment life cycle data in the cloud, obtaining the three-dimensional point cloud data therein, preprocessing the three-dimensional point cloud data, converting the preprocessed three-dimensional point cloud data into a triangular mesh surface model, and constructing a geometric model of the concrete production equipment using a surface fitting technology; Based on the geometric model of concrete production equipment, the finite element analysis method is used to perform numerical simulation and solve the stress state, vibration characteristics and heat conduction characteristics of the concrete production equipment, obtain the physical response data of the concrete production equipment under different working conditions, and build the physical model of the equipment; Obtain the condition monitoring parameters and performance parameters of concrete production equipment, use gated recurrent unit neural network to establish the nonlinear mapping relationship between condition monitoring parameters and performance parameters, and build an equipment degradation model for life prediction; The equipment geometric model, equipment physical model and equipment degradation model are integrated to construct a multi-level digital twin model of concrete production equipment.

4. The method according to claim 1, characterized in that: Based on the multi-level digital twin model, the pre-built equipment degradation model and data-driven model are integrated to build an equipment remaining life prediction model. The remaining life of the equipment is predicted using the equipment remaining life prediction model, including: Based on the multi-level digital twin model, a data-driven model is constructed. The data-driven model uses a convolutional neural network as the basic structure and adaptively extracts multi-scale degradation features related to the equipment degradation process from health monitoring data through a multi-scale feature extraction mechanism. Based on the obtained multi-scale degradation features, a joint loss function combining fault diagnosis loss and remaining life prediction loss is constructed, and the pre-constructed equipment degradation model is used as a regularization constraint term to train the data-driven model. The degradation trajectory output by the equipment degradation model is used as a priori constraint of the data-driven model to obtain prediction results that conform to the degradation trajectory, and the prediction results are used to feedback and optimize the equipment degradation model parameters. Repeat the iteration until the preset termination condition is met, and finally obtain a trained equipment remaining life prediction model. Based on the trained equipment remaining life prediction model, the fault diagnosis results are used as prior knowledge, and the Bayesian inference method is used to predict the remaining life of the components by combining physical priors with data-driven posteriori.

5. The method according to claim 4, characterized in that Based on the obtained multi-scale degradation features, the calculation formula of the joint loss function combining fault diagnosis loss and remaining life prediction loss is as follows: Where L represents the joint loss function, α represents the weight coefficient of the fault diagnosis model loss, N represents the number of devices, M represents the number of fault categories, and y ij Indicates the true label of whether the i-th device belongs to the j-th type of fault, represents the predicted value of the mean of the characteristic vector of the i-th device under the j-th type of fault, μ j represents the true value of the mean of the characteristic vector of the jth type of fault, β represents the weight coefficient of the remaining life prediction model loss, W p (·) represents the difference between the remaining life distributions, represents the predicted remaining life distribution of the ith device, P ri represents the true remaining life distribution, γ represents the weight coefficient of the regularization term, W represents the parameter matrix of the model, and ||·||2 represents the L2 norm.

6. The method according to claim 1, characterized in that According to the prediction results, the concrete production strategy is dynamically adjusted in the digital twin model to obtain the optimal concrete production scheduling plan, which is converted into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan, including: A multi-objective reinforcement learning optimization framework is constructed to model the concrete production process as a multi-objective Markov decision process, wherein the multi-objective Markov decision process includes a state space, an action space, a state transition probability matrix, a multi-objective reward function, a discount factor, and an initial state distribution, wherein the state space includes equipment state and process parameters, the action space includes equipment operation mode and production scheduling strategy, and the multi-objective reward function includes equipment remaining life, production efficiency, and energy consumption indicators; Based on the constructed multi-objective Markov decision process, a hierarchical target learning algorithm is designed, and a value function approximator and a policy function approximator are constructed for each optimization target of the multi-objective reward function. The value function and the policy function of each optimization target are estimated by using the value function approximator and the policy function approximator. According to the obtained value function estimation value and policy function estimation value, a target improvement operator is obtained, and the multi-objective reward function is updated based on the target improvement operator. Based on the updated multi-objective reward function, a non-dominated sorting genetic algorithm is used to search in the multi-objective Markov decision process to obtain an initial policy set. Based on the initial strategy set obtained, in the digital twin model, by building an interactive interface between the intelligent agent and the virtual environment, the Monte Carlo tree search algorithm is used to select the optimal exploration path through an adaptive exploration-utilization equilibrium strategy, generate large-scale optimization strategy sample data, and obtain the optimal strategy set; The optimal strategy set is input into the interactive decision recommendation system of the target learning algorithm, and the feasibility of the strategy is evaluated in combination with expert knowledge to generate the final optimal concrete production scheduling plan; The optimal concrete production scheduling plan is converted into equipment control instructions, and the automatic issuance and execution of the concrete production scheduling plan is realized through the interactive interface between the digital twin model and the actual production management system.

7. The method according to claim 6, characterized in that A value function approximator and a policy function approximator are constructed for each optimization objective of the multi-objective reward function. The value function and the policy function of each optimization objective are estimated using the value function approximator and the policy function approximator. The target improvement operator is obtained according to the obtained value function estimation value and policy function estimation value, and the calculation formula for updating the multi-objective reward function based on the target improvement operator is as follows: Among them, R′ k represents the updated multi-objective reward function, R k represents the initial multi-objective reward function, λ represents the learning rate, A represents the action space, δ represents the discount factor, S represents the state space, P(·) represents the state transition probability matrix, represents the value function of the kth optimization objective estimated by the value function approximator at state s′, represents the value function of the kth optimization objective estimated by the value function approximator at state s, ω represents the adjustment factor that balances the reward function and the policy function, represents the probability of the kth optimization objective choosing action a′ in state s estimated by the policy function approximator, represents the probability of selecting action a in state s for the kth optimization objective estimated by the policy function approximator.

8. A concrete production management system based on digital twin, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The first unit is used to collect health monitoring data of concrete production equipment in real time through multiple sensors, and collect full life cycle event data of concrete production equipment using IoT identification methods, and transmit the collected health monitoring data and full life cycle event data to the edge and cloud platforms respectively; The second unit is used to pre-process the health monitoring data using edge computing technology at the edge end, and extract the status characteristics of the equipment, perform real-time diagnosis based on the status characteristics of the equipment using a pre-trained fault prediction model, and update the annotation of the full life cycle event data according to the diagnosis results, and at the same time use the annotated full life cycle event data on the cloud platform to build a multi-level digital twin model of the equipment; The third unit is used to integrate the pre-built physical degradation model and data-driven model based on the multi-level digital twin model, build an equipment remaining life prediction model, use the equipment remaining life prediction model to predict the remaining life of the equipment, and dynamically adjust the concrete production strategy in the digital twin model according to the prediction results to obtain the optimal concrete production scheduling plan, and convert it into equipment control instructions to realize the automatic issuance and execution of the concrete production scheduling plan.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images