Concrete slump control system and method combining AI monitoring and ERP

Abstract

The application provides a concrete slump control system and method combining AI monitoring and ERP, and relates to the technical field of building materials.The system comprises a proportioning parameter acquisition module, which acquires concrete proportioning design parameters; a path determination module, which determines a transportation path according to a vehicle-mounted end and a mixing station end; an attenuation prediction module, which acquires vehicle-mounted operation condition data of the vehicle-mounted end to determine a transportation attenuation index; an out-machine prediction module, which acquires mixing images and inputs the images into a double-flow neural network to obtain a predicted out-machine slump; and a compensation control module, which transmits the transportation attenuation index to the mixing station end to perform compensation control on the predicted out-machine slump.The application solves the technical problem that, in the prior art, the influence of the transportation process on the concrete slump cannot be comprehensively considered, so that the concrete slump does not meet the design requirements when the concrete reaches the construction site, and achieves the technical effects of accurately controlling the concrete slump and improving the stability of the concrete quality.

Description

Technical Field

[0001] This application relates to the field of building materials technology, specifically to a concrete slump control system and method that combines AI monitoring and ERP. Background Technology

[0002] Concrete slump, as a key indicator of concrete workability and fluidity, has a significant impact on the final construction quality. Current methods for controlling concrete slump largely rely on preset parameters at the batching plant and the experience of operators. During transportation, losses are often estimated based on fixed times or distances, and adjustments are made using simple methods such as adding admixtures. These methods fail to fully consider the influence of various actual working conditions during transportation, such as the transportation route and the vehicle's operating status. This results in deviations in the slump of the concrete upon arrival at the construction site, making it difficult to meet design requirements and impacting construction quality and efficiency. Summary of the Invention

[0003] This application provides a concrete slump control system and method that combines AI monitoring and ERP, which solves the technical problem that the existing technology cannot comprehensively consider the influence of multiple factors such as transportation route and vehicle operation conditions on concrete slump, resulting in the concrete slump not meeting the design requirements when it arrives at the construction site. It achieves the technical effect of accurately controlling concrete slump and improving the stability of concrete quality.

[0004] In view of the above problems, on the one hand, this application provides a concrete slump control system combining AI monitoring and ERP. The system includes: a mix proportion parameter acquisition module, used to access the ERP work order system in the cloud platform to collect concrete mix proportion design parameters of the target work order; a path determination module, used to determine the transport vehicle end and the mixing plant end of the target work order, and determine the transport path according to the transport vehicle end and the mixing plant end; an attenuation prediction module, used to collect the vehicle operation condition data corresponding to the transport vehicle end according to the transport path, and perform concrete attenuation prediction based on the vehicle operation condition data to obtain the transport attenuation index of the transport path; an outlet prediction module, used to collect mixing images at the mixing plant end, input the mixing images into a dual-stream neural network to predict the concrete outlet slump, and obtain the predicted outlet slump; and a compensation control module, used to transmit the transport attenuation index to the mixing plant end to compensate and control the predicted outlet slump until the target slump in the concrete mix proportion design parameters is met.

[0005] On the other hand, this application also provides a concrete slump control method combining AI monitoring and ERP. The method includes: accessing the ERP work order system in a cloud platform to collect concrete mix design parameters of the target work order; determining the transport vehicle end and the mixing plant end of the target work order, and determining the transport route according to the transport vehicle end and the mixing plant end; collecting the vehicle operation condition data corresponding to the transport vehicle end according to the transport route, and performing concrete attenuation prediction based on the vehicle operation condition data to obtain the transport attenuation index of the transport route; collecting mixing images based on the mixing plant end, inputting the mixing images into a dual-stream neural network to predict the concrete discharge slump, and obtaining the predicted discharge slump; transmitting the transport attenuation index to the mixing plant end to compensate and control the predicted discharge slump until the target slump in the concrete mix design parameters is met.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] The mix proportioning parameter acquisition module accesses the ERP work order system from the cloud platform to collect concrete mix proportioning design parameters for the target work order, providing basic data for the entire scheme and ensuring clear target basis for subsequent control. The route determination module identifies the transport vehicle and mixing plant ends of the target work order and determines the transport route, laying the foundation for subsequent concrete state monitoring and prediction during specific transportation processes, allowing focus on specific transportation links. The attenuation prediction module collects on-vehicle operating condition data based on the transport route and performs concrete attenuation prediction to obtain a transport attenuation index, enabling early estimation of concrete slump loss during transportation and providing crucial predictive data for subsequent compensation control. The discharge prediction module acquires mixing images at the mixing plant and inputs the images into a dual neural flow network to predict the concrete discharge slump, accurately obtaining the actual slump at the production source for correlation analysis with attenuation during subsequent transportation. The compensation control module transmits the transport attenuation index to the mixing plant to compensate for the predicted discharge slump, achieving precise control of concrete slump through dynamic adjustments until the target slump is met, ensuring concrete quality meets design requirements.

[0008] In summary, this application achieves accurate prediction and dynamic compensation control of concrete slump during transportation. Through the synergistic effect of various modules, from collecting mix proportioning parameters to determining the transportation route, predicting transportation attenuation and slump at the discharge point, and finally implementing compensation control, a complete closed-loop control system is formed. This system ensures that the slump of the concrete arrives at the construction site within the design requirements, greatly improving the quality stability of the concrete, reducing construction quality problems caused by non-compliance with slump requirements, and enhancing the overall efficiency and reliability of concrete production and construction. It also plays a positive role in promoting the intelligent and precise development of the concrete industry.

[0009] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the structure of a concrete slump control system combining AI monitoring and ERP, provided in an embodiment of this application.

[0011] Figure 2 This is a flowchart illustrating the concrete slump control method combining AI monitoring and ERP, as provided in an embodiment of this application.

[0012] Explanation of reference numerals in the attached diagram: 10 for ratio parameter acquisition module, 20 for path determination module, 30 for attenuation prediction module, 40 for output prediction module, and 50 for compensation control module. Detailed Implementation

[0013] This application provides a concrete slump control system and method that combines AI monitoring and ERP. This solves the technical problem in the prior art where the slump of concrete does not meet the design requirements when it arrives at the construction site because the existing technology cannot comprehensively consider the influence of various factors such as transportation routes and vehicle operating conditions on the slump of concrete. This achieves the technical effect of accurately controlling the slump of concrete and improving the stability of concrete quality.

[0014] Example 1, as Figure 1 As shown in the embodiment of this application, a concrete slump control system combining AI monitoring and ERP is provided, the system comprising:

[0015] The mix proportioning parameter acquisition module 10 is used to access the ERP work order system in the cloud platform to collect the concrete mix proportioning design parameters of the target work order.

[0016] Specifically, the ERP work order system records concrete mix design parameters for each batch of concrete, including the formula, pouring location, quantity, required strength grade, and target slump. The module connects the work order data from the ERP system to the cloud platform via API or data interface, such as using a RESTful API or database middleware connection to achieve data synchronization. It automatically reads the mix design parameters corresponding to the target work order and uses them as benchmark data for subsequent prediction and control. For example, the ERP system generates a work order "Order No. 2025041801" with the following mix parameters: cement 350kg / m³, water 165kg / m³, sand 630kg / m³, aggregate 1200kg / m³, admixture 5kg / m³, and a target slump of 180mm. This information is synchronized to the cloud platform for use by subsequent modules.

[0017] The route determination module 20 is used to determine the transport vehicle end and the mixing plant end of the target work order, and determine the transport route according to the transport vehicle end and the mixing plant end.

[0018] Specifically, the "transport vehicle end" refers to the concrete mixer trucks carrying the concrete. The "batching plant end" refers to the concrete batching plant, which is usually the location where the concrete leaves the factory. Based on the locations of the concrete transport vehicles and the batching plant specified in the work order, the module calls GIS (Geographic Information System) and navigation services to determine the actual driving route of the transport vehicles from the batching plant to the construction site, i.e., the transport path. It records the route length, estimated time, and road conditions along the way, providing a reference for transport conditions to predict subsequent changes in concrete performance, so as to more accurately predict the decline in concrete slump.

[0019] The attenuation prediction module 30 is used to collect on-board operating condition data corresponding to the transport vehicle terminal according to the transport route, perform concrete attenuation prediction based on the on-board operating condition data, and obtain the transport attenuation index of the transport route.

[0020] Specifically, the module collects operational information of the transport vehicle, such as vehicle speed, vibration, and ambient temperature, through devices like the OBD vehicle terminal, gyroscope, and accelerometer, based on the transport route, obtaining onboard operating condition data. Using a pre-built transport attenuation prediction model, it predicts concrete attenuation based on this onboard operating condition data, forecasting the amount of slump reduction caused by vibration, time, and temperature, resulting in a transport attenuation index (i.e., slump reduction value), such as "ΔS=25mm". This index is used to quantitatively represent the changes in concrete performance during transport, providing accurate data support for subsequent compensation and adjustment.

[0021] The discharge prediction module 40 is used to acquire mixing images from the mixing plant, input the mixing images into a dual-stream neural network to predict the concrete discharge slump, and obtain the predicted discharge slump.

[0022] Specifically, high-definition industrial cameras and vibration sensors are installed at the mixing plant to capture images and collect vibration data of the concrete flowing out of the discharge port, obtaining mixing images. These images include video image sequences from multiple angles and vibration spectrograms. These images are then input into a pre-trained dual-stream neural network for predicting the concrete slump at discharge. The dual-stream neural network includes spatial and temporal streams, used to process image data and dynamically changing data, respectively. The spatial stream network of the dual-stream neural network performs feature analysis on the video image sequences, while the temporal stream network performs feature analysis on the vibration spectrograms, outputting image and temporal features. These features are then fused and predicted to obtain the predicted slump at discharge, i.e., the slump value of the concrete at discharge.

[0023] This module can make a relatively accurate prediction of the slump before the concrete leaves the site, which helps to identify potential problems in the mixing process in a timely manner and provides another important data source for subsequent compensation control, thereby improving the accuracy of the overall concrete slump control.

[0024] The compensation control module 50 is used to transmit the transportation attenuation index to the mixing plant to compensate and control the predicted slump at the discharge point until the target slump in the concrete mix design parameters is met.

[0025] Specifically, the module transmits the transportation attenuation index obtained from the attenuation prediction module 30 to the mixing plant. At the mixing plant, based on this index and the predicted slump at the discharge point obtained from the discharge prediction module 40, a feedback mechanism is used to adjust the water addition, extend the mixing time, or add admixtures to compensate for the slump until the target slump is met. The control logic can be automatically executed through a PLC system or a MES system. For example, if the transportation attenuation index indicates that the slump of the concrete will decrease by 20mm during transportation, while the predicted discharge slump is 180mm and the target slump is 200mm, then an appropriate amount of admixture needs to be added at the mixing plant to increase the slump. This module, through a prediction-based feedforward compensation mechanism, effectively avoids insufficient slump of the concrete arriving at the site, achieving closed-loop intelligent control throughout the entire process and improving project adaptability and concrete quality consistency.

[0026] Furthermore, the transportation route includes a first transportation route, wherein the first transportation route is the route for supplying concrete raw materials from the first transport vehicle to the mixing plant, and the route determination module 20 includes:

[0027] The first construction module is used to establish a raw material transportation attenuation model, which is obtained by training on historical operating condition data of the first transportation vehicle terminal.

[0028] The first prediction module is used to collect the first vehicle-mounted operating condition data of the first transport vehicle, call the raw material transport attenuation model to predict the attenuation of concrete raw materials on the first vehicle-mounted operating condition data, and output the attenuation index of transported raw materials.

[0029] The first compensation module is used to transmit the attenuation index of the transported raw materials to the mixing station for proportioning compensation control at the raw material stage.

[0030] Specifically, the first transport vehicle end refers to the transport vehicles and related equipment responsible for supplying concrete raw materials (such as cement, sand, and gravel) to the mixing plant. The first transport route refers to the transport path from the raw material trucks (such as vehicles transporting sand, gravel, and cement) to the mixing plant. Historical operating condition data of the first transport vehicle end is obtained from the database, including information such as vehicle speed, travel time, road conditions, and initial raw material data (raw material type and moisture content) during historical transportation. Historical raw material receiving records, such as raw material moisture content, are extracted from the raw material receiving records of the mixing plant for each historical operating condition record. This collected historical data is used to construct a training dataset. Supervised learning methods (such as XGBoost regression and support vector regression) are employed, using historical operating condition data as input and historical raw material receiving records as supervised output, to train a raw material transportation attenuation model to assess the loss or change in raw material performance during transportation.

[0031] During actual transportation, sensors and other equipment installed on the first transport vehicle collect the first vehicle's operating condition data. This data is then input into a pre-built raw material transportation attenuation model to predict the attenuation of concrete raw materials and output attenuation indicators, such as "sand moisture content decreased by 1.5%".

[0032] The transport material attenuation index obtained from the first prediction module is transmitted to the mixing plant. At the mixing plant, the concrete mix proportions are adjusted based on this index. For example, if the transport material attenuation index indicates a decrease in the moisture content of sand and gravel, the amount of water used during mixing is increased accordingly to ensure the final performance of the concrete. For instance, during transportation, cement raw materials experience temperatures as high as 35°C due to unsealed vehicles, resulting in some moisture evaporation; the model predicts a 0.8% reduction in water content. This compensation index is transmitted to the mixing plant, and the mixing plant control system automatically increases the planned water addition from 165 kg / m³ to 170 kg / m³, thereby ensuring that the concrete slump remains at the design target.

[0033] The above modules solve the problem of undetectable hidden quality changes during raw material transportation, realize feedforward compensation adjustment of the mix proportion before mixing, effectively improve the consistency of input quality of concrete mixtures, and enhance the stability and intelligent control level of the production stage.

[0034] Furthermore, the transportation route includes a second transportation route, which is the route for transporting the finished concrete from the mixing plant to the second transport vehicle. The route determination module 20 further includes:

[0035] The second construction module is used to establish a finished product transportation attenuation model, which is obtained by training on the historical operating condition data of the second transportation vehicle terminal and the slump difference label before and after transportation.

[0036] The second prediction module is used to collect the second vehicle-mounted operating condition data of the second transport vehicle, call the finished product transport attenuation model to predict the concrete slump attenuation of the second vehicle-mounted operating condition data, and output the transport slump attenuation index.

[0037] The second compensation module is used to transmit the transport slump attenuation index to the mixing plant for finished product compensation control during the discharge stage.

[0038] Specifically, the second transport vehicle end refers to the transport vehicles and related equipment responsible for transporting the finished concrete produced at the batching plant to the construction site. The second transport route refers to the transport route of the finished concrete from the batching plant to the construction site. The finished concrete transport attenuation model is a model established by combining vehicle operation data with slump as the output target, used to predict the slump decrease during transportation.

[0039] The second construction module collects historical operating data from the second transport vehicle and labels the difference in concrete slump before and after each transport. Then, a Long Short-Term Memory (LSTM) network is used for supervised learning on this data. The historical operating data from the second transport vehicle is used as the input vector, and the actual measured difference in slump before and after transport during historical transport is used as the regression label. Mean Squared Error (MSE) is used as the loss function, and the Adam optimizer is employed for weight training. The model parameters are continuously optimized in multiple iterations to make the predicted output as close as possible to the actual attenuation value, until the model converges, outputting the constructed finished product transport attenuation model.

[0040] The second prediction module collects second vehicle operating condition data during actual transportation of finished concrete, such as transportation time, ambient temperature, vehicle speed, and vehicle vibration amplitude. This data is then input into the finished concrete transportation attenuation model to predict and output the transportation slump attenuation index.

[0041] The second compensation module transmits the slump reduction index during transportation to the batching plant. At the batching plant, adjustments are made to the finished concrete about to leave the plant based on this index. This includes appropriately increasing the slump at the discharge point during concrete production or adjusting the dosage of admixtures, to ensure that the slump of the concrete still meets construction requirements after transportation to the construction site, thus achieving compensation control at the finished product stage. For example, if a truckload of concrete needs to be transported to a construction site 9 kilometers away from the batching plant in 30 minutes, the model predicts that transportation will cause a 28mm decrease in slump. Therefore, the batching plant control system changes the discharge slump from the initial 180mm to 208mm in advance.

[0042] The above modules enable feedforward prediction and dynamic compensation of the slump of finished products, effectively avoiding rework or waste at the construction site and ensuring construction quality and efficiency.

[0043] Furthermore, the second vehicle-mounted operating condition data includes transportation time, vehicle vibration amplitude, acceleration change, turning frequency, and ambient temperature; the finished product transportation attenuation model is used to extract transportation attenuation time-series features from the second vehicle-mounted operating condition data, predict slump attenuation based on the extracted transportation attenuation time-series features, and output transportation slump attenuation index.

[0044] Specifically, transportation time is the total time taken for finished concrete to travel from the batching plant to the construction site, measured in minutes or seconds. Vehicle vibration amplitude is the intensity of vibration in all directions (especially the vertical direction) during transportation, measured by accelerometers or gyroscopes. Acceleration variation indicates the speed fluctuation of the vehicle during transportation, mainly used to identify actions such as rapid acceleration and sudden braking. Turning frequency indicates the number of turns the vehicle makes per unit time, indirectly reflecting the degree of curvature change of the transportation path. Ambient temperature is the external air temperature during concrete transportation, directly affecting the rate of change in concrete slump.

[0045] The second prediction module collects real-time onboard operating condition data from sensors (such as accelerometers, gyroscopes, and thermometers) deployed on the transport vehicle. It then performs preprocessing such as noise filtering, data synchronization and timestamp correction, and outlier removal before inputting the data into the finished product transportation attenuation model to predict the slump attenuation of concrete.

[0046] The finished product transportation attenuation model adopts a structure design based on a long short-term memory network, possessing strong time-series dependency modeling capabilities. After receiving the second vehicle-mounted operating condition data, a time-series analysis is first performed. Using methods such as sliding window method, FFT, wavelet transform, and time-series statistical analysis, transportation attenuation time-series features from the original time-series data during the period from the transportation start point to the destination are extracted. These features include average vibration amplitude, vibration frequency spectral density, number of abrupt accelerations, average turning time interval, and slope of temperature change trend. Then, based on the extracted transportation attenuation time-series features, slump attenuation prediction is performed, outputting a transportation slump attenuation index. For example, in the model prediction, if a transportation batch experiences significant vibration during the 10-15 minute time interval, combined with the temperature rise trend at that time, the model predicts a slump loss of 10mm during this period. This prediction result will be used for precise compensation control of the slump at the mixing plant.

[0047] By introducing second vehicle-mounted operating condition data and extracting transportation attenuation time-series features, accurate modeling of complex dynamic conditions in concrete transportation is achieved, maintaining model adaptability under different climates, roads, and vehicle conditions.

[0048] Furthermore, the stirring image includes a sequence of video images from multiple angles and a vibration spectrum, and the outlet prediction module 40 is also used for:

[0049] The video image sequence and the vibration spectrum are input into the dual-stream neural network. Feature analysis is performed on the video image sequence based on the spatial flow network of the dual-stream neural network, and feature analysis is performed on the vibration spectrum based on the temporal flow network of the dual-stream neural network. Image features and temporal features are output. Feature fusion prediction is performed on the image features and the temporal features to obtain the predicted collapse degree.

[0050] Specifically, in order to accurately predict the slump of concrete leaving the mixer, the mixing images include a sequence of video images from multiple angles and corresponding vibration spectrum diagrams. The video images are used to intuitively reflect the appearance of the concrete during the mixing process, such as fluidity, boundary adhesion, and vortex formation. The vibration spectrum diagram is a frequency domain image generated by performing Fourier transform on the vibration data collected by the sensors during the mixing process, reflecting the dynamic energy changes during the operation of the mixing equipment.

[0051] The outflow prediction module 40 employs a dual-stream neural network structure, comprising a spatial stream network and a temporal stream network. The spatial stream network uses a multi-layer convolutional neural network (such as ResNet50) to process video image sequences from multiple angles, extracting spatial features between image frames, such as concrete color distribution, particle aggregation state, and edge diffusion. The temporal stream network takes the vibration spectrum generated by Fourier transform as input and uses a one-dimensional or two-dimensional convolutional neural network (such as 1D-CNN or the Inception module) to extract dynamic features such as spectral intensity distribution, energy band peak values, and dominant frequency variation trends. These spatial and temporal features are then fused using methods including fully connected concatenation, weighted averaging, and attention-based fusion. The fused feature vector is input to a regression prediction layer (such as an MLP or Transformer decoder), outputting the predicted outflow slump value. For example, if the image sequence shows a relatively smooth concrete surface and strong high-frequency components in the vibration spectrum, the model predicts that the current concrete is in a low slump state (e.g., 120mm); if the image shows a lot of tumbling of the concrete and significant low-frequency components in the spectrum, the predicted slump at the mixer is 180mm. This dual-stream neural network model enables the fusion perception and analysis of the visual and dynamic states during concrete mixing, thereby significantly improving the prediction accuracy of the slump at the mixer and providing a reliable basis for subsequent compensation control.

[0052] Furthermore, the compensation control module 50 includes:

[0053] The detection module is used to detect the real-time slump of the concrete at the second transport vehicle end.

[0054] The judgment module is used to determine whether the real-time slump of the concrete meets the target slump in the concrete mix design parameters. If the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters, a navigation command is sent to the second transport vehicle terminal to control the second transport vehicle terminal to return to the mixing plant terminal.

[0055] The third compensation module is used to control the concrete slump compensation at the mixing plant end according to the transportation attenuation index.

[0056] Specifically, to ensure that any abnormal slump in the concrete during transportation can be promptly remedied, the compensation control module 50 includes a detection module, a judgment module, and a third compensation module. The detection module is deployed on the second transport vehicle and is used to collect concrete slump parameters in real time and upload them to the cloud platform. For example, the current slump data can be obtained through an on-board automatic slump measurement device (such as an image processing-based drop distance detection device) and uploaded to the cloud platform.

[0057] After receiving the real-time slump data, the judgment module compares the real-time slump value with the target slump in the concrete mix design parameters of the ERP design work order to determine whether the real-time concrete slump meets the target slump, that is, whether the deviation between the two is within the allowable range. If it exceeds the allowable deviation (such as ±10mm), the return compensation control mechanism is triggered, and the vehicle communication system is used to send navigation instructions to the second transport vehicle to control the second transport vehicle to return to the mixing plant.

[0058] When the second transport vehicle returns to the mixing plant, the third compensation module automatically adjusts the batching ratio based on the previously recorded transport attenuation index (e.g., a loss of 20mm slump). This includes adding appropriate amounts of water-reducing agent, controlling the amount of water added, remixing, or making local adjustments to compensate for the slump loss and ensure that the concrete meets the original design requirements when it is discharged again.

[0059] Through the above modules, concrete that does not meet the requirements can be effectively adjusted, ensuring that even if performance fluctuations occur during dynamic transportation, closed-loop correction can be achieved, improving the stability of concrete quality, and ensuring that the performance of concrete meets the requirements of engineering design.

[0060] Furthermore, the navigation instructions are sent through the cloud platform, which includes multiple mixing plant terminals and multiple transport vehicle terminals. The judgment module includes:

[0061] The location acquisition module is used to acquire the real-time location of the second transport vehicle terminal based on the cloud platform if the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters.

[0062] The neighboring mixing plant end identification module is used to detect the relative distance between the second transport vehicle end and the multiple mixing plant ends based on the real-time location, and to obtain the neighboring mixing plant ends, wherein the neighboring mixing plant end is the mixing plant end with the smallest relative distance.

[0063] The instruction generation module is used to generate the navigation instructions based on the location of the nearest mixing plant.

[0064] Specifically, when the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters, the location acquisition module obtains the GPS positioning information of the current second transport vehicle terminal through the cloud platform, identifies the real-time location of the second transport vehicle terminal (represented in coordinate form), and feeds it back to the identification module of the neighboring mixing plant.

[0065] The neighboring batching plant identification module obtains the location information of multiple batching plants from the cloud platform. This location information is also in the form of geographic coordinates. For each batching plant, the distance between it and the second transport vehicle terminal is calculated using the Euclidean distance formula. By comparing the relative distances of multiple batching plants, the nearest batching plant is selected as the compensation station.

[0066] The instruction generation module generates navigation instructions based on the location of the nearest mixing plant and sends them to the onboard system of the transport vehicle through the cloud platform to guide its return journey.

[0067] Furthermore, the system described in this application embodiment also includes a training module, which is used for:

[0068] Initialize the dual-stream neural network; collect multiple mixing control parameter sample groups from the multiple mixing stations, perform transfer learning on the dual-stream neural network according to the multiple mixing control parameter sample groups respectively, and output multiple dual-stream neural networks; extract multiple network parameters of the multiple dual-stream neural networks, perform fusion learning on the multiple network parameters, and obtain updated multiple dual-stream neural networks; download the updated multiple dual-stream neural networks to the multiple mixing stations based on the cloud platform.

[0069] Specifically, the training module uses deep learning frameworks (such as TensorFlow or PyTorch) to create a two-stream neural network, initialize weight parameters, and prepare it for subsequent training.

[0070] To address the environmental differences among multiple batching plants (such as equipment model and mixer structure), mixing image sequences and vibration spectrum data, along with corresponding actual slump labels, are collected from the control systems or databases of multiple batching plants. This forms multiple sample groups for mixing control parameters, such as sample group A1 for site A and sample group B1 for site B. For each sample group, the mixing image sequence and vibration spectrum data are used as input data, and the corresponding mixing result (actual slump at discharge) is used as label data. Transfer learning is then performed on an initialized dual-stream neural network. During the transfer learning process, for each sample group from a different batching plant, some layers of the spatial and temporal stream networks are frozen (such as feature extraction layers, which typically learn general image and temporal features), and only the subsequent fully connected layers are retrained to adapt them to the concrete slump prediction task for that specific batching plant. In this way, the initial dual-stream neural network is individually adjusted using data from different batching plants, resulting in multiple dual-stream neural networks tailored to different batching plants.

[0071] The parameter weights (such as convolutional layer weights and LSTM gate weights) of each site's dual-stream neural network model are extracted. These parameters are then integrated using methods such as weighted averaging, attention mechanism fusion, or meta-learning strategies to obtain a more generalized set of network parameters. This set of parameters is then used to update the dual-stream neural network model for each site, resulting in multiple updated dual-stream neural networks. For example, the predictive performance of each network is evaluated on a validation set, its weight coefficients are calculated, and then the network parameters are weighted and averaged to obtain the fused network parameters. These fused network parameters are then used to update the original multiple dual-stream neural networks, enabling them to comprehensively consider the characteristics of multiple mixing stations and improving the model's generalization ability.

[0072] Based on the cloud platform, the updated dual-stream neural networks are downloaded to multiple mixing plants. The cloud platform then transmits the network model files to the equipment at each mixing plant via a data transmission protocol, achieving localized model deployment.

[0073] The training module trains and optimizes the dual-stream neural network through transfer learning and fusion learning, and distributes the updated network model to each batching plant, enabling the dual-stream neural network to adapt to the concrete production characteristics of different batching plants and improve prediction performance.

[0074] Furthermore, the compensation control module 50 is also used for:

[0075] After obtaining the predicted slump at the mixing plant, the actual slump at the mixing plant is obtained. The actual slump at the mixing plant is compared with the predicted slump at the mixing plant to calculate the prediction residual. The prediction residual is used as the error inverse input signal to optimize the parameters of the dual-stream neural network.

[0076] Specifically, the prediction residual refers to the difference between the actual slump at the mixer outlet and the model's predicted value, reflecting the accuracy of the model's prediction. After obtaining the predicted slump at the mixer outlet, the compensation control module 50 acquires the actual slump at the mixer outlet through a slump detection device installed at the outlet of the mixing plant. The actual value is compared with the predicted value, and the difference between the two is calculated, which is the prediction residual.

[0077] The prediction residuals are used as the backpropagation input signal to optimize the parameters of the two-stream neural network. The backpropagation algorithm calculates the gradient of the loss function (e.g., mean squared error) with respect to the weights and biases of each layer in the network. Based on this gradient information, the network parameters are adjusted to reduce the prediction residuals. For example, gradient descent can be used to update the network parameters according to a certain learning rate. The update formula is: New weights = Old weights - Learning rate × Gradient. This process is repeated continuously. As training data accumulates and the number of iterations increases, the prediction accuracy of the two-stream neural network gradually improves.

[0078] The compensation control module 50 uses the prediction residual to further optimize the dual-stream neural network, forming a closed-loop feedback control, continuously improving the prediction accuracy of the model and enhancing its adaptability and generalization ability to different operating conditions.

[0079] In summary, the concrete slump control system combining AI monitoring and ERP provided in this application has the following beneficial effects:

[0080] The mix proportioning parameter acquisition module 10 accesses the ERP work order system from the cloud platform to collect the concrete mix proportioning design parameters of the target work order, providing basic data for the entire scheme and ensuring that subsequent control has a clear target basis. The route determination module 20 determines the transport vehicle end and the mixing plant end of the target work order and determines the transport route, laying the foundation for subsequent concrete state monitoring and prediction during specific transportation processes, and can focus on specific transportation links. The attenuation prediction module 30 collects the corresponding vehicle operation condition data based on the transport route and performs concrete attenuation prediction to obtain the transportation attenuation index, which can predict the loss of concrete slump during transportation in advance, providing key predictive data for subsequent compensation control. The discharge prediction module 40 acquires mixing images at the mixing plant end and inputs the images into a dual neural flow network to predict the concrete discharge slump, which can accurately obtain the actual slump of concrete at the source of production, so as to perform correlation analysis with the attenuation during subsequent transportation. The compensation control module 50 transmits the transportation attenuation index to the mixing plant to compensate for the predicted slump at the discharge point. Through dynamic adjustment, it achieves precise control of the concrete slump until the target slump is met, ensuring that the concrete quality meets the design requirements.

[0081] In summary, this application achieves accurate prediction and dynamic compensation control of concrete slump during transportation. Through the synergistic effect of various modules, from collecting mix proportioning parameters to determining the transportation route, predicting transportation attenuation and slump at the discharge point, and finally implementing compensation control, a complete closed-loop control system is formed. This system ensures that the slump of the concrete arrives at the construction site within the design requirements, greatly improving the quality stability of the concrete, reducing construction quality problems caused by non-compliance with slump requirements, and enhancing the overall efficiency and reliability of concrete production and construction. It also plays a positive role in promoting the intelligent and precise development of the concrete industry.

[0082] Example 2, as Figure 2 As shown, based on the same inventive concept as in Embodiment 1 above, this application provides a method for controlling concrete slump by combining AI monitoring and ERP, the method comprising:

[0083] Step S1: Connect to the ERP work order system in the cloud platform to collect the concrete mix design parameters of the target work order.

[0084] Step S2: Determine the transport vehicle end and the mixing plant end of the target work order, and determine the transport route according to the transport vehicle end and the mixing plant end.

[0085] Step S3: Collect the on-board operating condition data corresponding to the transport vehicle terminal according to the transport route, and perform concrete attenuation prediction based on the on-board operating condition data to obtain the transport attenuation index of the transport route.

[0086] Step S4: Acquire mixing images at the mixing plant end, input the mixing images into a dual-stream neural network to predict the concrete slump at the mixer, and obtain the predicted slump at the mixer.

[0087] Step S5: Transmit the transport attenuation index to the mixing plant to compensate and control the predicted slump at the discharge point until the target slump in the concrete mix design parameters is met.

[0088] Furthermore, the transportation path includes a first transportation path, which is the path for supplying concrete raw materials from the first transport vehicle to the mixing plant; a raw material transportation attenuation model is established, which is obtained by training on historical operating condition data of the first transport vehicle; first vehicle operating condition data of the first transport vehicle is collected, and the raw material transportation attenuation model is called to predict the concrete raw material attenuation of the first vehicle operating condition data, and output the transportation raw material attenuation index; the transportation raw material attenuation index is transmitted to the mixing plant for proportion compensation control at the raw material stage.

[0089] Furthermore, the transportation path includes a second transportation path, which is the path for the second transport vehicle to transport the finished concrete from the mixing plant. A finished concrete transportation attenuation model is established, which is obtained by training on historical operating condition data of the second transport vehicle and slump difference labels before and after transportation. Second vehicle operating condition data of the second transport vehicle are collected, and the finished concrete transportation attenuation model is called to perform AI prediction of concrete slump attenuation on the second vehicle operating condition data, and outputs a transportation slump attenuation index. The transportation slump attenuation index is transmitted to the mixing plant for finished concrete compensation control during the discharge stage.

[0090] Furthermore, the finished product transportation attenuation model is invoked to predict the concrete slump attenuation of the second vehicle-mounted operating condition data, including:

[0091] The second vehicle-mounted operating condition data includes transportation time, vehicle vibration amplitude, acceleration change, turning frequency, and ambient temperature; the finished product transportation attenuation model is used to extract transportation attenuation time-series features from the second vehicle-mounted operating condition data, predict slump attenuation based on the extracted transportation attenuation time-series features, and output transportation slump attenuation index.

[0092] Furthermore, the mixing image is input into a dual-stream neural network to predict the concrete slump at the discharge point, resulting in a predicted slump, including:

[0093] The stirring images include a video image sequence from multiple angles and a vibration spectrum. The video image sequence and the vibration spectrum are input into the dual-stream neural network. Feature analysis is performed on the video image sequence based on the spatial flow network of the dual-stream neural network, and feature analysis is performed on the vibration spectrum based on the temporal flow network of the dual-stream neural network. Image features and temporal features are output. Feature fusion prediction is performed on the image features and the temporal features to obtain the predicted slump at the outlet.

[0094] Furthermore, after transmitting the transport attenuation index to the mixing plant to compensate for and control the predicted slump at the outlet, the process includes:

[0095] The real-time slump of the concrete at the second transport vehicle end is detected; it is determined whether the real-time slump of the concrete meets the target slump in the concrete mix design parameters; if the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters, a navigation command is sent to the second transport vehicle end to control the second transport vehicle end to return to the mixing plant end; the mixing plant end performs concrete slump compensation according to the transport attenuation index.

[0096] Furthermore, a navigation command is sent to the second transport vehicle terminal, wherein the navigation command is sent through the cloud platform, which includes multiple mixing plant terminals and multiple transport vehicle terminals; if the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters, the cloud platform obtains the real-time position of the second transport vehicle terminal; based on the real-time position, the relative distance between the second transport vehicle terminal and the multiple mixing plant terminals is detected, and the nearest mixing plant terminal is obtained, wherein the nearest mixing plant terminal is the mixing plant terminal with the smallest relative distance; the navigation command is generated based on the position of the nearest mixing plant terminal.

[0097] Furthermore, the cloud platform includes multiple mixing plant terminals, and the method further includes:

[0098] Initialize the dual-stream neural network; collect multiple sets of mixing control parameters from the multiple mixing stations, perform transfer learning on the dual-stream neural network according to the multiple sets of mixing control parameters, and output multiple dual-stream neural networks; extract multiple network parameters from the multiple dual-stream neural networks, perform fusion learning on the multiple network parameters, and obtain updated multiple dual-stream neural networks; the cloud platform downloads the updated multiple dual-stream neural networks to the multiple mixing stations accordingly.

[0099] Furthermore, after obtaining the predicted slump at the mixer, the actual slump at the mixer is obtained, and the error between the actual slump and the predicted slump is compared to calculate the prediction residual. The prediction residual is then used as the error inverse input signal to optimize the parameters of the dual-stream neural network.

[0100] Through the foregoing detailed description of the concrete slump control system combining AI monitoring and ERP, those skilled in the art can clearly understand that the concrete slump control method combining AI monitoring and ERP in this embodiment corresponds to the system disclosed in Embodiment 1, with corresponding execution steps and beneficial effects. For relevant details, please refer to the system section description.

[0101] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A concrete slump control system combining AI monitoring and ERP, characterized in that, The system includes: The mix proportioning parameter acquisition module is used to access the ERP work order system in the cloud platform to collect the concrete mix proportioning design parameters of the target work order; The route determination module is used to determine the transport vehicle end and the mixing plant end of the target work order, and determine the transport route according to the transport vehicle end and the mixing plant end; The attenuation prediction module is used to collect on-board operating condition data corresponding to the transport vehicle terminal according to the transport route, and to perform concrete attenuation prediction based on the on-board operating condition data to obtain the transport attenuation index of the transport route. Specifically, a pre-built transport attenuation prediction model is used to predict the slump attenuation caused by vibration, time and temperature based on these on-board operating condition data, and to obtain a transport attenuation index, namely the slump attenuation value. The outflow prediction module is used to acquire mixing images from the mixing plant, input the mixing images into a dual-stream neural network to predict the concrete outflow slump, and obtain the predicted outflow slump. The compensation control module is used to transmit the transportation attenuation index to the mixing plant to compensate for the predicted slump at the discharge point until the target slump in the concrete mix design parameters is met.

2. The concrete slump control system combining AI monitoring and ERP as described in claim 1, characterized in that, The transportation route includes a first transportation route, wherein the first transportation route is the route for supplying concrete raw materials from the first transport vehicle to the mixing plant, and the route determination module includes: The first construction module is used to establish a raw material transportation attenuation model, which is obtained by training on the historical operating condition data of the first transportation vehicle terminal. The first prediction module is used to collect the first vehicle operating condition data of the first transport vehicle terminal, call the raw material transport attenuation model to predict the concrete raw material attenuation of the first vehicle operating condition data, and output the transport raw material attenuation index. The first compensation module is used to transmit the attenuation index of the transported raw materials to the mixing station for proportioning compensation control at the raw material stage.

3. The concrete slump control system combining AI monitoring and ERP as described in claim 1, characterized in that, The transportation route includes a second transportation route, which is the route taken by the second transport vehicle to transport the finished concrete from the mixing plant. The route determination module further includes: The second construction module is used to establish a finished product transportation attenuation model, which is obtained by training on the historical operating condition data of the second transportation vehicle terminal and the slump difference label before and after transportation. The second prediction module is used to collect the second vehicle operating condition data of the second transport vehicle terminal, call the finished product transport attenuation model to predict the concrete slump attenuation of the second vehicle operating condition data, and output the transport slump attenuation index. The second compensation module is used to transmit the transport slump attenuation index to the mixing plant for finished product compensation control during the discharge stage.

4. The concrete slump control system combining AI monitoring and ERP as described in claim 3, characterized in that, The second type of vehicle operating condition data includes transportation time, vehicle vibration amplitude, acceleration change, turning frequency, and ambient temperature; The finished product transportation attenuation model is used to extract transportation attenuation time-series features from the second vehicle-mounted operating condition data, predict slump attenuation based on the extracted transportation attenuation time-series features, and output transportation slump attenuation index.

5. The concrete slump control system combining AI monitoring and ERP as described in claim 1, characterized in that, The stirring images include a sequence of video images from multiple angles and a vibration spectrum. The output prediction module is also used for: The video image sequence and the vibration spectrum are input into the dual-stream neural network. Feature analysis is performed on the video image sequence based on the spatial flow network of the dual-stream neural network, and feature analysis is performed on the vibration spectrum based on the temporal flow network of the dual-stream neural network. Image features and temporal features are output. The image features and the time features are fused and predicted to obtain the predicted collapse value.

6. The concrete slump control system combining AI monitoring and ERP as described in claim 3, characterized in that, The compensation control module includes: The detection module is used to detect the real-time slump of the concrete at the second transport vehicle end; The judgment module is used to determine whether the real-time slump of the concrete meets the target slump in the concrete mix design parameters. If the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters, a navigation command is sent to the second transport vehicle terminal to control the second transport vehicle terminal to return to the mixing plant terminal. The third compensation module is used to control the concrete slump compensation at the mixing plant end according to the transportation attenuation index.

7. The concrete slump control system combining AI monitoring and ERP as described in claim 6, characterized in that, The navigation instructions are sent through the cloud platform, which includes multiple mixing plant terminals and multiple transport vehicle terminals. The judgment module includes: The location acquisition module is used to acquire the real-time location of the second transport vehicle terminal based on the cloud platform if the real-time slump of the concrete does not meet the target slump in the concrete mix design parameters. The neighboring mixing plant end identification module is used to detect the relative distance between the second transport vehicle end and the multiple mixing plant ends based on the real-time location, and to obtain the neighboring mixing plant ends, wherein the neighboring mixing plant end is the mixing plant end with the smallest relative distance; The instruction generation module is used to generate the navigation instructions based on the location of the nearest mixing plant.

8. The concrete slump control system combining AI monitoring and ERP as described in claim 7, characterized in that, The system further includes a training module, which is used for: Initialize the two-stream neural network; Multiple sets of mixing control parameters are collected from the multiple mixing stations. The dual-stream neural network is then transferred and learned according to the multiple sets of mixing control parameters, and multiple dual-stream neural networks are output. Extract multiple network parameters from the multiple two-stream neural networks, perform fusion learning on the multiple network parameters, and obtain updated multiple two-stream neural networks; Based on the cloud platform, the updated dual-stream neural networks are downloaded to the corresponding mixing stations.

9. The concrete slump control system combining AI monitoring and ERP as described in claim 1, characterized in that, The compensation control module is also used for: After obtaining the predicted slump at the mixer, the actual slump at the mixer end is obtained, and the error between the actual slump and the predicted slump is compared to calculate the prediction residual. The prediction residual is used as the error inverse input signal to optimize the parameters of the two-stream neural network.

10. A method for controlling concrete slump by combining AI monitoring and ERP, characterized in that, The method is executed by the concrete slump control system combining AI monitoring and ERP as described in any one of claims 1-9, including: Integrate the ERP work order system into the cloud platform to collect concrete mix design parameters for the target work orders; Determine the vehicle-mounted end and the mixing plant end of the target work order, and determine the transportation route according to the vehicle-mounted end and the mixing plant end; Based on the transportation route, the vehicle-mounted terminal collects the corresponding vehicle-mounted operating condition data, and based on the vehicle-mounted operating condition data, concrete attenuation is predicted to obtain the transportation attenuation index of the transportation route. Based on the mixing image acquired at the mixing plant, the mixing image is input into a dual-stream neural network to predict the concrete slump at the discharge point, and the predicted slump at the discharge point is obtained. The transport attenuation index is transmitted to the mixing plant to compensate for and control the predicted slump at the discharge point until the target slump in the concrete mix design parameters is met.

Images