Method for predicting quality of ready-mixed concrete, quality prediction program, method for manufacturing ready-mixed concrete, and device for predicting quality of ready-mixed concrete

The method uses machine learning and neural networks to predict ready-mixed concrete quality accurately by inputting power load values and images, simplifying the process and improving accuracy in predicting slump, slump flow, and air content.

JP7603864B1Active Publication Date: 2024-12-20MITSUBISHI UBE CEMENT CORP

Patent Information

Application Number
JP2024053142
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2024-03-28
Publication Date
2024-12-20
Estimated Expiration
2044-03-28

AI Technical Summary

Technical Problem

Existing methods for predicting the quality of ready-mixed concrete using prediction models are complex and do not achieve high accuracy, particularly in determining qualities such as slump, slump flow, and air content.

Method used

A method and device that utilize machine learning, specifically neural networks, to predict the quality of ready-mixed concrete by inputting information such as power load values, vibrations, and images, without normalizing input data or deleting connections in fully connected layers, and using specific weight update formulas and loss functions to improve accuracy.

Benefits of technology

The method enhances the accuracy of predicting qualities like slump, slump flow, and air content in ready-mixed concrete while simplifying the prediction process, ensuring the concrete meets target quality specifications.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007603864000001_ABST
    Figure 0007603864000001_ABST
Patent Text Reader

Abstract

To improve the accuracy of quality prediction while avoiding complication of work involved in quality prediction when performing quality prediction using a prediction model. [Solution] A method for predicting the quality of fresh concrete, comprising: an acquisition step of acquiring input information including first information relating to at least one of the power load value of a mixer that produces fresh concrete by mixing concrete materials or mixes the fresh concrete, vibrations caused by the falling or flowing down of fresh concrete, and an image of fresh concrete, and second information relating to a target quality of the fresh concrete when it is used; and a prediction step of predicting the quality of the fresh concrete based on a prediction model previously constructed by machine learning to output quality information indicating the quality of the fresh concrete in response to the input of the input information, and the input information acquired in the acquisition step.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present disclosure relates to a quality prediction method for ready-mix concrete, a quality prediction program, and a quality prediction device for ready-mix concrete. [Background technology]

[0002] Patent Document 1 discloses a method for predicting the quality of ready-mix concrete using a prediction model. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-124304 Summary of the Invention [Problem to be solved by the invention]

[0004] The present disclosure provides a quality prediction method, a quality prediction program, and a quality prediction device for ready-mixed concrete that can improve the accuracy of quality prediction while avoiding the complication of work associated with quality prediction when performing quality prediction using a prediction model. [Means for solving the problem]

[0005] [1] A method for predicting the quality of fresh concrete, comprising: an acquisition step of acquiring input information including first information relating to at least one of the power load value of a mixer that produces fresh concrete by mixing concrete materials or that mixes the fresh concrete, vibrations caused by the falling or flowing down of the fresh concrete, and an image of the fresh concrete, and second information relating to a target quality of the fresh concrete when it is used; and a prediction step of predicting the quality of the fresh concrete based on a prediction model previously constructed by machine learning so as to output quality information indicating the quality of the fresh concrete in response to the input of the input information, and the input information acquired in the acquisition step.

[0006] [2] The method for predicting quality of ready-mixed concrete described in [1] above, wherein the quality of ready-mixed concrete includes one or more qualities of slump, slump flow, and air content.

[0007] [3] The quality of fresh concrete includes two or more qualities of slump, slump flow, and air content, and the prediction model is constructed to output the two or more qualities in response to the input of the input information. The method for predicting quality of fresh concrete described in [2] above.

[0008] [4] A method for predicting quality of ready-mixed concrete described in any one of [1] to [3] above, wherein the prediction model is constructed by machine learning using a neural network that does not perform an operation to normalize the input information and an operation to cause some connections in the fully connected layer to be missing.

[0009] [5] A method for predicting the quality of ready-mixed concrete described in any one of [1] to [4] above, wherein the second information includes information indicating the target quality of the ready-mixed concrete at the time of shipment, in addition to information indicating the target quality of the ready-mixed concrete at the time of use.

[0010] [6] The method for predicting quality of ready-mixed concrete described in any one of [1] to [5] above, wherein the predictive model is constructed by machine learning using a neural network, and in the machine learning using a neural network when constructing the predictive model, one selected from the group consisting of Adam, AdamBelief, Adamax, AdaBound, Adagrad, AMSGRAD, AMSBound, RMSprop, SgdW, Momentum, and Nesterov is used as a weight update formula.

[0011] [7] The method for predicting quality of ready-mixed concrete described in any one of [1] to [6] above, wherein the prediction model is constructed by machine learning using a neural network, and in the machine learning using the neural network when constructing the prediction model, a function is used as a loss function such that the loss value L(a) is always less than or equal to the absolute value of the error a.

[0012] [8] A quality prediction program for causing a computer to execute the quality prediction method for ready-mix concrete according to any one of [1] to [7] above.

[0013] [9] A quality prediction device for fresh concrete, comprising: an input information acquisition unit that acquires input information including first information relating to at least one of the power load value of a mixer that produces fresh concrete by mixing concrete materials or mixes the fresh concrete, vibrations caused by the falling or flowing down of the fresh concrete, and an image of the fresh concrete, and second information relating to a target quality of the fresh concrete when it is used; a prediction model that is pre-constructed by machine learning to output quality information that indicates the quality of the fresh concrete in response to input of the input information; and a prediction calculation unit that predicts the quality of the fresh concrete based on the input information acquired by the input information acquisition unit. Effect of the Invention

[0014] According to the present disclosure, there is provided a quality prediction method for ready-mixed concrete, a quality prediction program, and a quality prediction device for ready-mixed concrete that can improve the accuracy of quality prediction while avoiding the complication of work associated with quality prediction when performing quality prediction using a prediction model. [Brief description of the drawings]

[0015] [Figure 1] FIG. 1 is a schematic diagram showing an example of a ready-mix concrete manufacturing system. [Diagram 2] FIG. 2 is a block diagram illustrating an example of a functional configuration of the control device. [Diagram 3] FIG. 3 is a diagram illustrating data relating to power load values. [Figure 4] FIG. 4 is a block diagram illustrating an example of a hardware configuration of the control device. [Diagram 5] Fig. 5(a) is a diagram showing an example of a process flow in the learning phase, and Fig. 5(b) is a diagram showing an example of a process flow in the evaluation phase. [Figure 6] 6(a) and 6(b) are schematic diagrams showing an example of a calculation process using a prediction model, and Fig. 6(c) is a schematic diagram showing an example of a calculation process using a neural network. [Figure 7] FIG. 7 is a graph showing an example of a loss function. [Figure 8] FIG. 8 is a graph showing an example of a comparison result between the predicted value and the correct value. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Hereinafter, an embodiment will be described with reference to the drawings. In the description, the same elements or elements having the same functions are given the same reference numerals, and duplicated explanations will be omitted. Fig. 1 shows a schematic diagram of a ready-mix concrete manufacturing system equipped with a quality prediction device according to an embodiment.

[0017] [Ready-mix concrete manufacturing system] First, an overview of a ready-mix concrete manufacturing system will be described. The manufacturing system 1 shown in Fig. 1 is a system for manufacturing ready-mix concrete. The manufacturing system 1 kneads concrete materials to manufacture ready-mix concrete.

[0018] The concrete materials used in the manufacturing system 1 include cement, admixtures, coarse aggregates, fine aggregates, water, and admixtures. Examples of the coarse aggregates include gravel, crushed stone, slag coarse aggregates, lightweight coarse aggregates, recycled coarse aggregates, recovered aggregates, and coarse aggregates made by mixing these. The gravel includes mountain gravel, land gravel, river gravel, and sea gravel. The slag coarse aggregates include blast furnace slag aggregates, ferro-nickel slag aggregates, electric furnace oxidized slag aggregates, and coal gasification slag aggregates. The lightweight coarse aggregates include natural lightweight aggregates, by-product lightweight aggregates, and artificial lightweight aggregates. The coarse aggregates may include crushed rock, or limestone crushed stone.

[0019] Examples of fine aggregates include sand, crushed sand, slag fine aggregate, lightweight fine aggregate, recycled fine aggregate, recovered aggregate, and fine aggregates made by mixing these. Sand includes mountain sand, land sand, river sand, sea sand, etc. Slag fine aggregates include blast furnace slag aggregate, ferro-nickel slag aggregate, copper slag aggregate, electric furnace oxidized slag aggregate, coal gasification slag aggregate, etc. Lightweight fine aggregates include natural lightweight aggregate, by-product lightweight aggregate, artificial lightweight aggregate, etc.

[0020] Rock types of the crushed stone and crushed sand include, for example, igneous rocks, sedimentary rocks, metamorphic rocks, quartzite, limestone, domalite, or peridotite. Igneous rocks include granite, diorite, gabbro, porphyrite, diabase, rhyolite, andesite, basalt, or serpentinite. Sedimentary rocks include conglomerate, sandstone, shale, slate, or tuff. Metamorphic rocks include gneiss or schist.

[0021] The manufacturing system 1 loads the manufactured ready-mix concrete onto the transport vehicle 200. After the ready-mix concrete is loaded onto the transport vehicle 200, the transport vehicle 200 transports the ready-mix concrete to a site (e.g., a construction site) where the ready-mix concrete is to be used. Examples of the transport vehicle 200 include an agitator vehicle (mixer vehicle) or a dump truck. The manufacturing system 1 may manufacture ready-mix concrete from concrete materials so as to satisfy a target quality (required quality) set for each site. For example, an operator of the manufacturing system 1 determines a mix of concrete materials so as to satisfy the target quality set for each site, and inputs an operation instruction to the manufacturing system 1.

[0022] The target quality set for each site is different from the quality of the ready-mixed concrete produced by the production system 1 before it is shipped (at the time of shipment). For example, in the production system 1, the quality of the ready-mixed concrete produced by the production system 1 before it is shipped is controlled (inspected, etc.) so that the target quality set for each site is met. In this disclosure, the target quality set for each site is referred to as the "target quality when ready-mixed concrete is used." The time when ready-mixed concrete is used corresponds to the time when ready-mixed concrete is received at the site. The target quality when ready-mixed concrete is used is set, for example, as shown in (1) to (3) below.

[0023] (1) Japan Industrial Standards (National Standards) In one example, the target quality of the ready-mixed concrete when used includes a target value of at least one quality of slump, slump flow, and air content. The on-site target values ​​of slump, slump flow, and air content may be specified by a purchaser or an orderer according to "Table 1 - Types and Classifications of Ready-Mixed Concrete" in "JIS A 5308:2019 (Ready-Mixed Concrete)" or "JIS A 5308:2024 (Ready-Mixed Concrete)". (2) Building Standards Act (Law) In addition to the target quality stipulated in the above JIS, target quality may be set for building materials designated as building materials that are subject to certification by the Minister of Land, Infrastructure, Transport and Tourism in Article 37, Paragraph 1, Item 2 of the Building Standards Act. Ready-mixed concrete that does not conform to JIS can be used in buildings if it is certified by the Minister of Land, Infrastructure, Transport and Tourism. This means that ready-mixed concrete with target qualities not stipulated in the above JIS (for example, slump 23 cm, slump flow 65 cm, etc.) can be used in buildings (building foundations, main structures, parts important for safety, fire prevention, or hygiene, etc.). (3) Standards established by academic societies In cases where there is no mention in the above-mentioned JIS, the target quality when using ready mixed concrete may be set based on technical standards established by each academic society, such as the "Standard Specifications for Concrete" established by the Japan Society of Civil Engineers, a public interest incorporated association, and the "Standard Specifications for Construction Works and Commentary JASS5 Reinforced Concrete Works" established by the Architectural Institute of Japan, a general incorporated association. In addition, "JIS A 5308:2019 (Ready Mixed Concrete)" or "JIS A 5308:2024 (Ready Mixed Concrete)" described in (1) corresponds to "Japanese Industrial Standards designated by the Minister of Land, Infrastructure, Transport and Tourism" as specified in Article 37, Paragraph 1, Item 1 of the Building Standards Act. In this sense, the second information on the target quality at the time of use of ready mixed concrete may be the target quality of the designated building materials as specified in Article 37, Paragraph 1, Item 1 of the Building Standards Act. Alternatively, it may be the target quality of the designated building materials that are subject to certification by the Minister of Land, Infrastructure, Transport and Tourism as specified in Article 37, Paragraph 1, Item 2 of the same Act as described in (2). In addition, it may be the target quality set by standards set by academic societies such as "Standard Specifications for Concrete" or "Standard Specifications for Building Construction and Commentary JASS5 Reinforced Concrete Construction" as described in (3).

[0024] In order to manage the quality of ready-mixed concrete before it is shipped, a target quality of ready-mixed concrete at the time of shipment may be determined based on the target quality of ready-mixed concrete at the time of use. The target quality of ready-mixed concrete at the time of shipment may be set according to in-house standards or the like determined for each factory. In setting the target quality of ready-mixed concrete at the time of shipment, at least one piece of information may be taken into consideration: the state (quality) of the materials used during production, the season (temperature), the type of concrete, the target quality at the time of use, and the transportation time. The target quality of ready-mixed concrete at the time of shipment is set, for example, by adding a value determined by the in-house standards to the target quality of ready-mixed concrete at the time of use.

[0025] The manufacturing system 1 includes, for example, a manufacturing apparatus 100 and a control device 10. The manufacturing apparatus 100 is an apparatus that manufactures ready-mix concrete based on operation instructions from the control device 10. The manufacturing apparatus 100 manufactures ready-mix concrete by mixing concrete materials. The manufacturing apparatus 100 includes, for example, a material storage area 101, a transport device 104, a storage bottle 111, a measuring bottle 112, a collection hopper 113, a mixer 114, and a loading hopper 115.

[0026] The material yard 101 is a place where concrete materials are stored. The material yard 101 includes a plurality of silos 102. The plurality of silos 102 are containers that store at least a portion of the concrete materials by material type. The plurality of silos 102 include, for example, a silo 102 that stores coarse aggregate, a silo 102 that stores fine aggregate, and a silo 102 that stores cement.

[0027] The transporting device 104 is a device that transports the concrete materials stored in the multiple silos 102 to the storage bottle 111. The transporting device 104 includes, for example, a belt conveyor that transports the concrete materials. The transporting device 104 may transport the concrete materials by type at different times. In one example, based on an operation instruction from the control device 10, a specific material from among the various concrete materials is transferred to the transporting device 104 and transported to the storage bottle 111.

[0028] Storage bottle 111 temporarily stores various types of concrete materials. Various types of concrete materials are transported (conveyed) to storage bottle 111 from material storage yard 101 by transport device 104. Storage bottle 111 is configured to store various types of concrete materials individually. Hereinafter, "concrete material" may be simply referred to as "material". Various materials stored in storage bottle 111 are supplied to measuring bottle 112 as necessary.

[0029] The measuring bottle 112 is disposed below the storage bottle 111. The measuring bottle 112 operates based on an operation instruction from the control device 10, and individually measures various materials. When the measuring bottle 112 detects a target amount of material instructed by the control device 10, it supplies the material to the collecting hopper 113. When water is supplied to the measuring bottle 112, an admixture may be mixed into the water. The collecting hopper 113 is disposed below the measuring bottle 112. The collecting hopper 113 collects various materials discharged from the measuring bottle 112, and supplies the collected various materials to the mixer 114. Note that the manufacturing apparatus 100 does not need to be provided with the collecting hopper 113, and various materials may be supplied to the mixer 114 from the measuring bottle 112.

[0030] The mixer 114 is disposed below the collecting hopper 113. The mixer 114 is a device that mixes concrete materials. The mixer 114 produces ready-mixed concrete by mixing (kneading) aggregate, cement, water, admixtures, and the like. That is, the mixer 114 produces ready-mixed concrete by mixing the concrete materials. The ready-mixed concrete is discharged from the bottom of the mixer 114 into the loading hopper 115. The mixer 114 may be a tilting mixer, a horizontal single-shaft mixer, a horizontal double-shaft mixer, or a pan-type mixer. The mixer 114 includes, for example, two stirring members 114a and a mixer drive unit 114b.

[0031] The stirring member 114a is a member that stirs various materials supplied to the mixer 114. The two stirring members 114a are arranged side by side inside the main body (container) of the mixer 114 and are provided so as to be freely rotatable. Each of the two stirring members 114a includes a rotating shaft that extends in one horizontal direction. The mixer driving unit 114b rotates the rotating shafts of the two stirring members 114a based on an operation instruction from the control device 10. The mixer driving unit 114b includes, for example, a driving source such as a motor that imparts driving force to the stirring members 114a. An opening and closing port is provided at the bottom of the main body of the mixer 114 for discharging the manufactured ready-mix concrete into the loading hopper 115.

[0032] The loading hopper 115 is disposed below the mixer 114 and temporarily stores the ready-mixed concrete. The loading hopper 115 supplies the temporarily stored ready-mixed concrete to the transport vehicle 200.

[0033] The manufacturing apparatus 100 described above is an example of a ready-mix concrete manufacturing apparatus, and the ready-mix concrete manufacturing apparatus may be configured in any manner as long as it is capable of mixing concrete materials using a mixer and producing ready-mix concrete.

[0034] In this disclosure, a unit of ready-mixed concrete produced by one mixing in the mixer 114 and loaded onto the transport vehicle 200 is defined as "one batch." The process executed by the production system 1 for producing one batch of ready-mixed concrete is defined as "batch processing." In one example, one to three batches of ready-mixed concrete are loaded onto one transport vehicle 200. For example, when two batches of ready-mixed concrete are loaded onto one transport vehicle 200, two batch processes according to the same production conditions are performed at different times (in different orders).

[0035] <Control device (quality prediction device)> The control device 10 is a device that controls the manufacturing equipment 100. The control device 10 is configured with one or more computers. When the control device 10 is configured with multiple computers, these computers are connected to each other so that they can communicate with each other. The control device 10 controls the manufacturing equipment 100 according to set operating conditions. At least a part of the operating conditions may be determined by instructions from an operator such as a worker.

[0036] An input device 12 and a monitor 14 may be connected to the control device 10. The input device 12 is a device that inputs information indicating instructions from an operator or the like to the control device 10. The input device 12 may be any device capable of inputting desired information, and may be a keyboard (keypad), an operation panel, or a mouse. The monitor 14 is a device for displaying information from the control device 10 to an operator or the like. The monitor 14 may be any device capable of displaying graphics, and may be a liquid crystal display. The input device 12 and the monitor 14 may be integrated together, such as a touch panel. The control device 10, the input device 12, and the monitor 14 may be integrated together, such as a tablet computer (tablet terminal).

[0037] The control device 10 may have a function of predicting the quality of ready-mixed concrete manufactured by the manufacturing apparatus 100, in addition to controlling the manufacturing apparatus 100. In this case, the control device 10 constitutes a quality prediction device that predicts the quality of ready-mixed concrete (a device for predicting quality of ready-mixed concrete). In the following description, the quality of ready-mixed concrete predicted by the control device 10 is the quality after it has been manufactured by the manufacturing apparatus 100 and before it is shipped to the site (i.e., the quality of the ready-mixed concrete at the time of shipment).

[0038] The quality of fresh concrete to be predicted by the control device 10 may include one or more qualities of slump, slump flow, and air volume. The quality of fresh concrete to be predicted by the control device 10 may include two or more qualities of slump, slump flow, and air volume. The quality of fresh concrete to be predicted by the control device 10 may be one quality of slump, slump flow, and air volume, or may be two or more qualities of slump, slump flow, and air volume.

[0039] The control device 10 is configured to execute at least an acquisition step and a prediction step. The acquisition step is a step of acquiring input information including information related to the power load value of the mixer 114 and information related to the target quality at the time of use of the ready-mixed concrete whose quality is to be predicted (prediction target). The prediction step is a step of predicting the quality of the ready-mixed concrete to be predicted based on a prediction model previously constructed by machine learning so as to output quality information indicating the quality of the ready-mixed concrete in response to the input of the input information, and the input information acquired in the acquisition step.

[0040] 2 shows an example of functional components (hereinafter referred to as "functional blocks") included in the control device 10. The control device 10 has, for example, as functional blocks, an operation control unit 22, an operation information acquisition unit 24, a target quality information acquisition unit 26, a model construction unit 30, a model holding unit 32, a prediction calculation unit 28, and a display output unit 34. The processes executed by these functional blocks correspond to the processes executed by the control device 10.

[0041] The operation control unit 22 controls the manufacturing apparatus 100 to manufacture ready-mixed concrete according to predetermined operating conditions. At least a part of the operating conditions may be determined by an operator such as a worker each time ready-mixed concrete is manufactured. The operation control unit 22 may control the mixer driving unit 114b of the mixer 114 so that the rotation speed of the mixer driving unit 114b of the mixer 114 follows a target rotation speed defined in the operating conditions. When controlling the mixer driving unit 114b, the operation control unit 22 may adjust the power (e.g., current value) supplied to the mixer driving unit 114b. If the ready-mixed concrete to be manufactured is hard, the power load value tends to be large, and if the ready-mixed concrete to be manufactured is soft, the power load value tends to be small.

[0042] The operation information acquisition unit 24 acquires information related to the operation when the ready-mixed concrete is manufactured. The operation information acquisition unit 24 acquires information related to the power load value of the mixer 114 (hereinafter referred to as "first information"). The power load value of the mixer 114 may be a value indicating the power (kW) itself supplied to the mixer 114, or may be a value indicating the current value (A) supplied to the mixer 114. Alternatively, the power load value of the mixer 114 may be replaced with a value indicating the load oil pressure (MPa). The information related to the power load value of the mixer 114 may be continuous time series data while the mixer 114 is operating during the processing of one batch, or may be statistical data obtained from the time series data.

[0043] FIG. 3 shows a schematic diagram of time series data related to the power load value of the mixer 114 in one batch of processing. The time series data related to the power load value is data obtained by repeatedly measuring the power (kW) supplied to the mixer 114 at a predetermined sampling period, for example. The operation information acquisition unit 24 may calculate (acquire) at least one of the following statistical data: a fluctuation range (difference between the maximum value and the minimum value), a fall width (difference between the maximum value and the final value), a total value within an arbitrarily set time, an average value, a standard deviation, a coefficient of variation, a median, a first quartile, a third quartile, kurtosis, and skewness. The operation information acquisition unit 24 may calculate at least one of a minimum value and a maximum value as well as at least one of the above-mentioned groups as statistical data. The statistical data acquired by the operation information acquisition unit 24 may include a statistical value obtained by calculation using two or more power load values ​​included in the time series data.

[0044] In the graph shown in FIG. 3, "P1" represents an initial value in the time series data. The initial value P1 is the power load value at the start of mixing in the time series data. "P2" represents the minimum value in the time series data. The minimum value P2 is the minimum power load value after the time when the initial value P1 is obtained. Note that the power load value at the start of mixing (initial value P1) may also be the minimum. "P3" represents the maximum value in the time series data. The maximum value P3 is the maximum power load value after the time when the initial value P1 is obtained. "P4" represents the final value in the time series data. The final value P4 is the power load value at the time when the operation control unit 22 determines that the condition for ending mixing by the mixer 114 is satisfied. For example, when a predetermined time has elapsed since the start of driving the stirring member 114a, the operation control unit 22 determines that the above condition is satisfied and stops driving the stirring member 114a.

[0045] The target quality information acquisition unit 26 acquires information related to the target quality of the ready-mixed concrete when it is used (hereinafter referred to as "second information"). When the quality of the target predicted by the control device 10 includes slump, the target quality information acquisition unit 26 may acquire information indicating a target value of slump when the ready-mixed concrete is used as the second information. When the quality of the target predicted by the control device 10 includes slump flow, the target quality information acquisition unit 26 may acquire information indicating a target value of slump flow when the ready-mixed concrete is used as the second information. When the quality of the target predicted by the control device 10 includes air content, the target quality information acquisition unit 26 may acquire information indicating a target value of air content when the ready-mixed concrete is used as the second information. The target quality information acquisition unit 26 may further acquire information indicating a target quality at the time of shipment of the ready-mixed concrete as the second information. That is, the second information may include information indicating the target quality at the time of shipment of the ready-mixed concrete in addition to information indicating the target quality at the time of use of the ready-mixed concrete.

[0046] The target quality information acquisition unit 26 acquires the second information, for example, based on an input from an operator, such as a worker, via the input device 12. In one example, the target quality information acquisition unit 26 acquires the second information together with operation conditions related to the production of the ready-mix concrete to be produced. The target quality information acquisition unit 26 may acquire the second information based on an input from an operator, such as a worker, via the input device 12 at the timing when the control device 10 performs quality prediction. The operation information acquisition unit 24 and the target quality information acquisition unit 26 described above function as an input information acquisition unit that acquires input information including the first information and the second information.

[0047] The model construction unit 30 constructs a model for predicting the quality of ready-mixed concrete (hereinafter, referred to as "prediction model M"). The prediction model M is a model that outputs quality information (quality value) indicating the quality of ready-mixed concrete in response to input of input information including the first information and the second information. The model construction unit 30 constructs the prediction model M by machine learning based on the input information and the correct quality value associated with the input information. The prediction model M may be constructed to output, as the quality information, one or more predicted values ​​of a predicted value of slump, a predicted value of slump flow, and a predicted value of air volume. The prediction model M may be constructed to output, as the quality information, a predicted value of the ratio of slump flow to slump (slump flow / slump) instead of or in addition to the one or more predicted values.

[0048] Machine learning refers to a method in which a machine (computer) autonomously finds laws or rules by repeatedly learning based on given information. The prediction model M can be constructed using an algorithm and a data structure. The prediction model M is realized, for example, using a neural network, which is an information processing model that mimics the mechanism of the human brain. There is no particular limitation on the specific algorithm of the machine learning performed when constructing the prediction model M. The neural network has, for example, an input layer, one or more intermediate layers, and an output layer. By including one or more intermediate layers, a more complex prediction model M can be constructed, and prediction accuracy can be improved.

[0049] The model construction unit 30 may autonomously construct a prediction model M for predicting the quality of ready-mix concrete by performing machine learning using data provided as input for machine learning and correct answer data (correct answer values ​​for slump, etc.) that are the output of the machine learning. The input for the machine learning is various data sets of input information including the first information and the second information. The output of the machine learning is data (numerical values) that indicate the quality of the ready-mix concrete. The model construction unit 30 iteratively learns a model that outputs predicted values ​​for slump, etc., using multiple combinations of the data sets of input information and correct answer values ​​for slump, etc.

[0050] The stage of autonomously constructing the prediction model M corresponds to a learning phase. The learning phase may be performed before the production phase in which ready-mix concrete is manufactured, or may be performed at an early stage of the production phase. The model holding unit 32 holds the prediction model M constructed by the model construction unit 30. The prediction model M, which is a trained model, may be transferable between computers. Therefore, the prediction model M constructed in the control device 10 may be used in another manufacturing system different from the manufacturing system 1, and the model holding unit 32 may hold the prediction model M constructed in the other manufacturing system.

[0051] In the evaluation phase, the prediction calculation unit 28 predicts the quality of the fresh concrete to be predicted based on the input information acquired by the operation information acquisition unit 24 and the target quality information acquisition unit 26 (input information acquisition unit) and the prediction model M. The prediction calculation unit 28 inputs the acquired input information to the prediction model M and acquires a predicted value output from the prediction model M. The input information acquired in the evaluation phase (input information for evaluation) is information regarding the unknown quality of the fresh concrete.

[0052] The display output unit 34 outputs information (quality value) indicating the quality predicted by the prediction calculation unit 28 to the monitor 14. As a result, the predicted value of quality is displayed on the monitor 14, and an operator such as a worker can grasp the predicted value of the quality of the ready-mix concrete (manufactured ready-mix concrete) that is the prediction target.

[0053] As shown in Fig. 4, the control device 10 includes a circuit 50. The circuit 50 includes a processor 51, a memory 52, a storage 53, and an input / output port 54. The storage 53 is configured with one or more non-volatile memory devices such as a flash memory or a hard disk. The storage 53 stores at least a quality prediction program that causes a computer to execute the acquisition step and the prediction step. The storage 53 stores a quality prediction program for configuring each functional block of the control device 10.

[0054] The memory 52 is composed of one or more volatile memory devices such as a random access memory. The memory 52 temporarily stores a quality prediction program loaded from the storage 53. The processor 51 is composed of one or more arithmetic devices such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit). The processor 51 configures each functional block of the control device 10 by executing the quality prediction program loaded into the memory 52. ​​The calculation results by the processor 51 are temporarily stored in the memory 52. ​​The input / output port 54 inputs and outputs information between the input device 12, the monitor 14, the mixer 114, etc. in response to a request from the processor 51.

[0055] The timer 55 measures the elapsed time by, for example, counting a reference pulse at a constant period. The circuit 50 is not necessarily limited to one in which each function is configured by a program. For example, the circuit 50 may be configured with at least some of its functions by a dedicated logic circuit or an ASIC (Application Specific Integrated Circuit) that integrates the dedicated logic circuit. The quality prediction program may be provided in a state in which it is fixedly recorded on a tangible recording medium such as a CD-ROM, a DVD-ROM, or a semiconductor memory. Alternatively, the quality prediction program may be provided via a communication network as a data signal superimposed on a carrier wave.

[0056] [Method of manufacturing ready-mix concrete] Next, an example of a method for producing ready-mix concrete executed in the production system 1 will be described. The method for producing ready-mix concrete includes a production process and a quality prediction process. The production process is a process for producing ready-mix concrete. The quality prediction process is a process for predicting the quality of the ready-mix concrete produced in the production process. The quality prediction process may be executed in a period that overlaps with at least a portion of the period during which the production process is repeatedly executed.

[0057] The manufacturing process includes, for example, a transporting process, a weighing process, a feeding process, a mixing process, a discharging process, and a loading process. In the transporting process, various types of concrete materials are transported to storage bottles 111 by a transport device 104, and the various materials are individually supplied to the storage bottles 111. In the weighing process, various materials are individually supplied from the storage bottles 111 to weighing bottles 112, and the various materials are weighed in the weighing bottles 112. In the weighing process, when the measured amount of each material reaches a predetermined set amount, the material is discharged to a collecting hopper 113. In the feeding process, after all types of materials are collected in the collecting hopper 113, the materials in the collecting hopper 113 are fed (supplied) to a mixer 114.

[0058] In the mixing process, multiple types of concrete materials are mixed in the mixer 114. In the mixing process, the control device 10 may control the mixer driving unit 114b according to predetermined operating conditions. In the mixing process, the power supplied from the control device 10 to the mixer driving unit 114b may be adjusted so that the rotation speed of the mixer driving unit 114b follows a target rotation speed.

[0059] In the discharging process, after mixing of the concrete materials is completed in the mixer 114, the ready-mixed concrete is discharged from the mixer 114 to the loading hopper 115. In the loading process, the ready-mixed concrete discharged into the loading hopper 115 is loaded onto the transport vehicle 200.

[0060] The quality prediction process (quality prediction method) includes a model construction process in a learning phase and a quality evaluation process in an evaluation phase. In the quality prediction process, the model construction process is executed before the quality evaluation process. An example of the model construction process and an example of the quality evaluation process will be described below.

[0061] (Model building process) 5(a) is a flowchart showing an example of a series of processes executed in the model construction process. This model construction process is executed before the above-mentioned manufacturing process is executed in the manufacturing apparatus 100, or at an early stage when the above-mentioned manufacturing process is started. In this model construction process, for example, ready-mix concrete actually manufactured in the manufacturing apparatus 100 is used as ready-mix concrete for learning.

[0062] In the model construction process, step S11 is executed first. In step S11, for example, an operator such as a worker prepares learning data for machine learning. The learning data is composed of a plurality of data sets. Each of the plurality of data sets includes input information (learning input information) including the first information and the second information obtained when the learning ready mixed concrete is manufactured, and a correct answer value of quality information (for example, slump) associated with the input information.

[0063] The correct value of the quality information may be a value obtained by actually measuring the quality of the ready-mixed concrete for learning. In one example, after the ready-mixed concrete for learning is loaded onto the transport vehicle 200, a part of the ready-mixed concrete is extracted by a worker or the like. Then, the slump, slump flow, air content, and the like of the extracted ready-mixed concrete are measured by the worker or the like, and at least a part of these actual measured values ​​is used as the correct value in the learning data.

[0064] Next, step S12 is executed. In step S12, for example, the model construction unit 30 of the control device 10 performs machine learning using the learning data (multiple data sets) prepared in step S11 to construct a prediction model M. The model construction unit 30 may construct the prediction model M by machine learning using a neural network.

[0065] Each of Fig. 6(a) and Fig. 6(b) shows a schematic diagram of a prediction model M constructed by the model construction unit 30. The prediction model M shown in Fig. 6(a) is a model that outputs a single quality prediction value in response to input of input information. The prediction model M shown in Fig. 6(b) is a model that outputs two or more types of quality prediction values ​​in response to input of input information. In the present disclosure, machine learning performed so that the prediction model M outputs a single quality prediction value (only a prediction value of one type of quality) is referred to as "single-task learning". Moreover, machine learning performed so that the prediction model M outputs prediction values ​​of two or more types of quality is referred to as "multi-task learning".

[0066] The model construction unit 30 may construct the prediction model M by performing single-task learning. In this case, the model construction unit 30 may construct a prediction model M for each type of quality. For example, the model construction unit 30 constructs two or more models among a prediction model M that outputs a predicted value of a slump, a prediction model M that outputs a predicted value of a slump flow, and a prediction model M that outputs a predicted value of an air volume. The model construction unit 30 may construct the prediction model M by performing multi-task learning. For example, the model construction unit 30 constructs a prediction model M (one prediction model M) that outputs a predicted value of each of two or more qualities among a slump, a slump flow, and an air volume. The model construction unit 30 may construct a prediction model M (one prediction model M) that outputs a ratio of a slump flow to a slump in addition to one or more qualities among a slump, a slump flow, and an air volume.

[0067] An example of the prediction model M will be described below using simplified formulas for ease of understanding. The prediction model M constructed by the model construction unit 30 can be simply expressed as, for example, the following formulas (1) and (2).

number

number

[0068] In formula (2), Y represents an output value of quality, and in a prediction model M constructed by single-task learning, it is an output value of a single quality. In a prediction model M constructed by multi-task learning, Y is an output value of any one of two or more types of quality, and formulas (1) and (2) are calculated for each type of quality. N is an integer of two or more and represents the number of input data. x represents various input values ​​included in the input data, and the input data corresponds to the above-mentioned input information, and the input data includes at least a value related to the power load value included in the first information (e.g., two or more statistics) and a value indicating the target quality included in the second information.

[0069] wi is a weight (coefficient), and b is a bias term (coefficient). f(U) represents the activation function. The activation function is a linear function (identity function) or a nonlinear function such as a polynomial, absolute value, step function, sigmoid function, hardsigmoid function, logsigmoid function, softmax function, logsoftmax function, softmin function, softplus function, softsign function, tanh function, tanhShrink function, hardtanh function, tanhexp function, ReLU function, ReLU6 function, Leaky-ReLU function, PReLU function, ELU function, SELU function, CELU function, Swith function, Mish function, ACON function, etc.

[0070] The model construction unit 30 may determine the weight wi and bias term b in formula (1) so as to minimize the error by repeatedly evaluating the error and loss value between Y (predicted value) obtained by formula (2) and the correct value of quality using learning data. The model construction unit 30 may use any type of loss function as a function for evaluating the error between Y (output value from an intermediate model at an intermediate stage of constructing the prediction model M) obtained by formula (2) and the correct value of quality. For example, the model construction unit 30 may use one selected from the group consisting of Huber loss function (HuberLoss), Mean Absolute Error (MAE), and ε-insensitive loss function (ε-insensitiveloss) as the loss function.

[0071] The Huber loss function can be expressed, for example, as in the following formula (3). The ε-allowable loss function can be expressed, for example, as in the following formula (4). L(a) is the loss value, and a is the error between the predicted value and the correct value. δ and ε are parameters that can be set arbitrarily, and are set, for example, to 1.0.

number

number

[0072] FIG. 7 illustrates various loss functions. Specifically, FIG. 7 illustrates the relationship between the error a and the loss value L(a) in the mean squared error (MSE) in addition to the relationship between the error a and the loss value L(a) in the Huber loss function (δ=1.0), the mean absolute error, and the ε-allowable loss function (ε=0.5). In the graph of FIG. 7, the horizontal axis is the error a, and the vertical axis is the loss value L(a). In the mean squared error, when the error a exceeds 1, the loss value L(a) becomes larger than the absolute value of the error a, whereas in the other three loss functions, the relationship that the loss value L(a) is equal to or smaller than the absolute value of the error a holds regardless of the value of the error a. From the viewpoint of improving the prediction accuracy, it is preferable to use a function that always makes the loss value L(a) equal to or smaller than the absolute value of the error a as the loss function. The mean absolute error or the ε-allowable loss function may be used as the function that always makes the loss value L(a) equal to or smaller than the absolute value of the error a. Alternatively, by setting δ of the Huber loss function to 1.0 or less, it is possible to use it as a function in which the loss value L(a) is always equal to or less than the absolute value of the error a.

[0073] The model construction unit 30 may repeatedly update the weights wi by the gradient method so as to minimize the error and loss value evaluated by the loss function. The model construction unit 30 may use any type of update formula (weight update formula) when updating the weights wi. For example, the model construction unit 30 uses one selected from the group consisting of Adam, AdamBelief, Adamax, AdaBound, Adagrad, AMSGRAD, AMSBound, RMSprop, SgdW, Momentum, and Nesterov as the weight update formula. The weight update formula is also called an optimization algorithm or an optimization method.

[0074] The model construction unit 30 may construct the prediction model M by machine learning using a neural network that does not perform an operation to normalize the input information and an operation to cause some connections in the fully connected layer to be missing. The operation to normalize the input information means an operation (batch normalization) that normalizes (standardizes) each value included in the input information that is input to the prediction model M so that the average is 0 and the standard deviation is 1.

[0075] The operation of missing some connections in the fully connected layer means that, when performing machine learning using a neural network, learning is performed while deactivating an arbitrary proportion of nodes (Dropout). The operation process of the prediction model M when some connections in the fully connected layer are missing (when Dropout is performed) is shown in FIG. 6(c). The model construction unit 30 may construct the prediction model M by machine learning using a neural network that performs at least one of an operation to normalize input information and an operation to missing some connections in the fully connected layer.

[0076] 5(a), after step S12 is executed, step S13 is executed. In step S13, for example, the model holding unit 32 stores the prediction model M built in step S12. This completes the model building process.

[0077] (Quality evaluation process) 5(b) is a flowchart showing an example of a series of processes executed in the quality evaluation process. This quality evaluation process is executed, for example, during a period overlapping at least a part of the period during which the manufacturing process is executed by the manufacturing apparatus 100.

[0078] In the quality evaluation process, first, the control device 10 executes step S21. In step S21, for example, the control device 10 waits until it is time to evaluate the quality of the ready-mixed concrete to be evaluated. The ready-mixed concrete to be evaluated is also the ready-mixed concrete to be manufactured by the manufacturing apparatus 100. The evaluation timing may be predetermined to a certain time period in a day, or may be predetermined to a timing to execute a certain number of batch processes in a day. The evaluation timing may be a timing at which an instruction to execute the evaluation is received from an operator such as a worker.

[0079] Next, the control device 10 executes step S22. In step S22, for example, the operation information acquisition unit 24 and the target quality information acquisition unit 26 acquire input information including the first information (information related to the power load value) and the second information (information related to the target quality) when the ready mixed concrete to be evaluated was produced. The input information acquired in step S22 is input information for evaluation in which quality information (for example, slump) is unknown.

[0080] Next, the control device 10 executes step S23. In step S23, for example, the prediction calculation unit 28 predicts quality information of the ready-mix concrete to be evaluated based on the input information acquired in step S22 and the prediction model M held in the model holding unit 32. In one example, the prediction calculation unit 28 inputs the input information acquired in step S22 to the prediction model M, and acquires a predicted value of one or more qualities of slump, slump flow, and air content output from the prediction model M.

[0081] Next, the control device 10 executes step S24. In step S24, for example, the display output unit 34 displays the predicted value of quality acquired in step S23 on the monitor 14. This allows an operator such as a worker to check the predicted value of quality.

[0082] This completes the quality evaluation process. The control device 10 may execute a series of processes from steps S21 to S24 each time one batch of ready-mixed concrete is produced (each time a batch process is performed). The control device 10 may execute a series of processes from steps S21 to S24 each time multiple batches of ready-mixed concrete are produced (each time a multiple batch process is performed).

[0083] [Variations] The series of processes shown in Figures 5(a) and 5(b) are examples and can be modified as appropriate. In the series of processes, one step and the next step may be executed in parallel, or some steps may be executed in an order different from the example described above. Instead of at least some of the steps of the series of processes, or in addition to the series of processes, steps with contents different from the example described above may be executed.

[0084] The first information may include at least one of an initial value, a minimum value, a maximum value, and a final value in the time-series data related to the power load value. The input information acquired by the control device 10 (input information acquisition unit) and input to the prediction model M may include information related to mixing by the mixer 114 (information other than the power load value). Examples of the information related to mixing include the mixing amount of one batch of concrete, the mixing time, the time when the power load value reaches the maximum value, and the time from when the power load value reaches the maximum value until the fresh concrete is discharged from the mixer 114. The input information input to the prediction model M may include the nominal strength, information indicating the type of cement, information indicating the type of admixture, information indicating the type of fine aggregate, information indicating the type of coarse aggregate, and information indicating the amount of additive added.

[0085] The first information included in the input information that is the input of the prediction model M may be information related to vibration caused by the falling or flowing down of fresh concrete, instead of the power load value of the mixer 114. The information related to vibration may be information related to vibration caused by the falling or flowing down of fresh concrete from the mixer 114. As shown in FIG. 1, the manufacturing system 1 may include a vibration sensor 18.

[0086] The vibration sensor 18 is a sensor that detects vibrations (vibration magnitude) caused by the falling or flowing down of the fresh concrete produced by the mixer 114. The vibrations caused by the falling or flowing down include vibrations caused by both the falling and flowing down of the fresh concrete. The vibration sensor 18 may be capable of detecting the magnitude of vibrations in two or more directions caused by the falling or flowing down of the fresh concrete. The vibration sensor 18 is attached to the loading hopper 115, for example, to detect the magnitude of vibration of the loading hopper 115. The vibration sensor 18 may be installed on any side wall of the loading hopper 115. The fresh concrete discharged from the mixer 114 may fall against the side wall on which the vibration sensor 18 is installed.

[0087] The vibration sensor 18 may be of any type as long as it is capable of detecting the magnitude of vibration of the loading hopper 115. The vibration sensor 18 is, for example, a sensor that detects the acceleration of the loading hopper 115. The vibration sensor 18 may detect acceleration in each of two directions along the side wall of the loading hopper 115 that are perpendicular to each other, and acceleration in a direction perpendicular to the side wall of the loading hopper 115. The vibration sensor 18 may detect acceleration in each direction at a predetermined sampling period. The vibration sensor 18 outputs information indicating the detection result to the control device 10.

[0088] The information relating to vibration among the input information to be input to the prediction model M may be time series data of detection values ​​by the vibration sensor 18 obtained during a period including the time point when the fresh concrete is discharged into the loading hopper 115 (hereinafter referred to as the "discharge period"). The information relating to vibration may be a statistic obtained from the time series data instead of or in addition to the time series data during the discharge period.

[0089] The first information included in the input information that is the input of the prediction model M may be information relating to an image of ready-mixed concrete instead of the power load value of the mixer 114. The information relating to the image of ready-mixed concrete is information relating to an image obtained by capturing an image of ready-mixed concrete being manufactured in the manufacturing apparatus 100. As shown in FIG. 1, the manufacturing system 1 may include an image sensor 16.

[0090] The image sensor 16 is a sensor (camera) that captures an image of the ready-mixed concrete being manufactured. In the present disclosure, capturing an image of the ready-mixed concrete being manufactured in the manufacturing apparatus 100 includes capturing an image of the ready-mixed concrete being manufactured in the mixer 114, in addition to capturing an image of the ready-mixed concrete after it has been manufactured in the mixer 114. Capturing an image of the ready-mixed concrete after it has been manufactured in the mixer 114 includes capturing an image of the ready-mixed concrete in the loading hopper 115, and capturing an image of the ready-mixed concrete being poured from the loading hopper 115 into the transport vehicle 200. The image data obtained by the imaging of the image sensor 16 may be either still image data or video data. The image sensor 16 may perform imaging based on an operation instruction from the control device 10. The image sensor 16 outputs the image data obtained by imaging to the control device 10.

[0091] The image sensor 16 may be arranged so as to be able to capture an image of the ready-mixed concrete being manufactured in the mixer 114. The image sensor 16 may be arranged so as to be able to capture an image of the ready-mixed concrete stored in the space in the loading hopper 115. When there is a possibility that the ready-mixed concrete will rebound in the mixer 114 or the loading hopper 115, the image sensor 16 may be arranged at a position where the image is less susceptible to the influence of the rebounded ready-mixed concrete. The image sensor 16 may be arranged so as to be able to capture an image of the ready-mixed concrete being discharged from the loading hopper 115. The manufacturing system 1 may include two or more image sensors 16 provided at two or more different locations.

[0092] In addition to the prediction of the quality of the ready-mixed concrete by the control device 10, the quality of the ready-mixed concrete may be measured periodically (e.g., several times a day) in the manufacturing apparatus 100. In this case, the prediction model M may be updated based on the actual measurement value of the quality of the ready-mixed concrete and input information at the time when the actual measurement value was obtained (e.g., information related to the power load value, or an image of the ready-mixed concrete). In the prediction step, the quality of the ready-mixed concrete may be predicted by the updated prediction model M. Note that even when the updated prediction model M is used, the process of predicting the quality of the ready-mixed concrete is still performed based on the prediction model M and the input information acquired in the acquisition step.

[0093] The first information acquired by the input information acquisition unit of the control device 10 and included in the input information to be input to the prediction model M may be information relating to two or more of the vibration caused by the falling or flowing down of fresh concrete from the mixer 114, an image of the fresh concrete, and the power load value of the mixer 114. As described above, the first information acquired by the input information acquisition unit of the control device 10 and included in the input information to be input to the prediction model M may be information relating to at least one of the vibration caused by the falling or flowing down of fresh concrete from the mixer 114, an image of the fresh concrete, and the power load value of the mixer 114.

[0094] The manufacturing system 1 may include, as a functional block, a quality prediction device (quality prediction device for ready-mix concrete) having at least an operation information acquisition unit 24, a target quality information acquisition unit 26, a prediction calculation unit 28, and a model storage unit 32, in addition to the control device 10. A computer constituting the quality prediction device may be communicably connected to the control device 10. Hereinafter, the quality prediction device having at least the operation information acquisition unit 24, the target quality information acquisition unit 26, the prediction calculation unit 28, and the model storage unit 32 will be simply referred to as the "quality prediction device."

[0095] In the above example, the control device 10 functioning as a quality prediction device predicts the quality of ready-mixed concrete after it is manufactured by the manufacturing device 100 and before it is shipped to the site. The timing at which the quality is predicted by the quality prediction device is not limited to this example. The quality prediction device may predict the quality of ready-mixed concrete during transportation (including upon arrival at the site) to the site where the ready-mixed concrete is used. For example, the transport vehicle 200 is provided with a mixer 204 that mixes the ready-mixed concrete (see FIG. 1), and the input information acquisition unit of the quality prediction device acquires information related to the power load value of the mixer 204 as at least a part of the first information. The mixer 204 in the transport vehicle 200 such as an agitator vehicle is also called a drum. As described above, at least a part of the first information may be information related to the production of ready-mixed concrete by mixing concrete materials, or the power load value of the mixer (114, 204) that mixes the ready-mixed concrete after production.

[0096] In the above example, the ready-mixed concrete is manufactured at a location (manufacturing apparatus 100) different from the site. The location where the ready-mixed concrete is manufactured is not limited to this example. Concrete materials (for example, concrete materials excluding water) may be transported to the site by a transport vehicle, and ready-mixed concrete may be manufactured at the site. At that time, water may be added to the concrete materials transported by the transport vehicle by a mixer provided on the transport vehicle, and the materials may be mixed to manufacture ready-mixed concrete. When ready-mixed concrete is manufactured at the site, the quality prediction device may predict the quality of the ready-mixed concrete at the site. The input information acquisition unit of the quality prediction device may acquire, as at least a part of the first information, information related to the power load value of the mixer provided on the transport vehicle when ready-mixed concrete is manufactured at the site.

[0097] In one example among the various examples described above, at least a part of the matters described in the other examples may be combined.

[0098] [Verification of prediction results using the prediction model] Next, we will explain the results of verifying the prediction results of quality information using a prediction model M that inputs input information including the first information and the second information, using a data set in which the correct answer value is known.

[0099] (Impact of target quality) The effect of the second information, which is information indicating the target quality when fresh concrete is used, was verified. In this verification, 2765 training data sets were prepared to construct a prediction model by machine learning, and 789 evaluation data sets, in which the correct answer value is known and for evaluation, were prepared. In the training data set and the evaluation data set, various input values ​​included in the input information are associated with the correct answer value of the quality information. As the first information of the input information in the training data set and the evaluation data set, statistics obtained from time series data related to the power load value are used. Using the 2765 training data sets, a prediction model M was constructed using input information including the first information and the second information. A prediction model M was constructed by multitask learning to output slump, slump flow, and air volume.

[0100] In order to verify the influence of the second information on the prediction accuracy, a comparative model was constructed using the same 2765 learning datasets after removing the second information from the input information. The influence of the second information on the prediction accuracy was evaluated for each of slump, slump flow, and air volume. Using 789 evaluation datasets, the prediction results of the model and the correct answer value in the evaluation dataset were compared for each of the prediction model M and the comparative model. The case where the prediction result of the model is included in the range obtained by adding a predetermined tolerance to the correct answer value in the evaluation dataset is defined as the correct answer, and the accuracy rate was used as the evaluation index. Figure 8 shows a graph showing the results of verifying the influence of the second information.

[0101] Regarding the verification results for slump in FIG. 8, when the tolerance is "±0.5 cm", the data set in which the prediction result (predicted value) by the model is within the range of ±0.5 cm of the correct value is regarded as the correct answer, and the percentage of the data sets judged to be correct is shown as the accuracy rate (%). The accuracy rates (%) are also calculated in the same manner when the tolerance is "±1.0 cm", "±1.5 cm", "±2.0 cm", and "±2.5 cm". "Yes" indicates the evaluation result by the prediction model M including the second information (target quality) in the input information, and "No" indicates the evaluation result by the comparative model not including the second information in the input information. It can be seen that the accuracy rate (%) when predicting slump using the prediction model M is higher than when predicting slump using the comparative model.

[0102] As for the verification results of slump flow in Fig. 8, the accuracy rate (%) was calculated for each tolerance, changing it to "±2.5 cm", "±3.0 cm", "±5.0 cm", and "±7.5 cm", in the same way as the verification of slump flow. As for slump flow, it can be seen that the accuracy rate (%) when predicting slump flow using prediction model M is higher than when predicting slump flow using the comparison model.

[0103] As with the verification of slump, the verification results for the air volume in Fig. 8 show the accuracy rate (%) for each tolerance, which was changed to "±0.3%, "±0.5%, "±1.0%, and "±1.5%". As for the air volume, it can be seen that the accuracy rate (%) when predicting the air volume using prediction model M is higher than when predicting the air volume using the comparative model.

[0104] (Effect of Batch Normalization Dropout) When constructing a model for predicting the quality information of fresh concrete, the effects of the operation (Batch Normalization) that normalizes input information and the operation (Dropout) that causes some data to be missing in the fully connected layer were verified. A prediction model M was constructed while changing the presence or absence of these operations and the missing rate when causing some data to be missing, and the accuracy rate (%) of the allowable discrimination for each of slump, slump flow, and air volume was calculated. In the verification, the above-mentioned 2765 learning datasets and the above-mentioned 789 evaluation datasets were used. The accuracy rate (%) of the allowable discrimination for slump is shown in Table 1 below, the accuracy rate (%) of the allowable discrimination for slump flow is shown in Table 2 below, and the accuracy rate (%) of the allowable discrimination for air volume is shown in Table 3 below.

[0105] [Table 1]

[0106] [Table 2]

[0107] [Table 3]

[0108] In the conditions of Tables 1 to 3, "BN" stands for Batch Normalization (a normalizing operation), "DO" stands for Dropout (an operation that causes some connections to be missing), and "P" stands for the missing rate in Dropout. "BN, DO not used" indicates that predictive model M was constructed by training with a neural network that does not include Batch Normalization and Dropout. A column with only "BN" indicates that predictive model M was constructed by training with a neural network that includes Batch Normalization and does not include Dropout. A column with only "DO" indicates that predictive model M was constructed by training with a neural network that includes Dropout but does not include Batch Normalization. A column with "BN+DO" indicates that predictive model M was constructed by training with a neural network that includes Batch Normalization and Dropout.

[0109] From the results shown in Tables 1 and 2, it can be seen that for slump and slump flow, there is no significant difference in the accuracy rate (%) depending on whether or not batch normalization and dropout are used. From the results shown in Table 3, it can be seen that for air volume, the accuracy rate (%) tends to be higher when batch normalization and dropout are not used compared to when at least one of batch normalization and dropout is used.

[0110] (Impact of the type of weight update formula) The effect of the type of weight update equation used when constructing a model for predicting quality information of fresh concrete was verified. A prediction model M was constructed while changing the type of weight update equation, and the accuracy rate (%) of the allowable discrimination for the air content was calculated. In the verification, the above-mentioned 2765 training datasets and the above-mentioned 789 evaluation datasets were used. The verification results are shown in Tables 4 and 5 below.

[0111] [Table 4]

[0112] [Table 5]

[0113] In Table 5, "(x 0.01)" indicates that each value included in the input information in the learning dataset was multiplied by 0.01 so that each value fell within the range of -1 to 1, and then machine learning was performed. From the results shown in Tables 4 and 5, it can be seen that when one selected from the group consisting of Adam, AdamBelief, Adamax, AdaBound, Adagrad, AMSGRAD, AMSBound, RMSprop, SgdW, Momentum, and Nesterov is used as the weight update formula, the accuracy rate (%) is higher than when AdamW, Lars, Adadelta, and Sgd are used. Although the data is omitted, the accuracy rate (%) was calculated for each update formula shown in Tables 4 and 5 for slump and slump flow, and when Lars and Sgd were used, the accuracy rate (%) was lower than when other update formulas were used.

[0114] (Influence of loss function type) The effect of the type of loss function used when constructing a model for predicting the quality information of fresh concrete was verified. A prediction model M was constructed while changing the type of loss function, and the accuracy rate (%) of the allowable discrimination was calculated for each of slump, slump flow, and air content. As the Huber loss function, the function shown in the above formula (3) was used, and the value of δ was set to 1.0. As the ε-allowable loss function, the function shown in the above formula (4) was used, and the value of ε was set to 1.0. In the verification, the above-mentioned 2765 training data sets and the above-mentioned 789 evaluation data sets were used. The verification results are shown in Tables 6, 7, and 8 below.

[0115] [Table 6]

[0116] [Table 7]

[0117] [Table 8]

[0118] From the results shown in Tables 6, 7, and 8, it can be seen that when one selected from the group consisting of the Huber loss function, the mean absolute error, and the ε-tolerable loss function is used as the loss function, the accuracy rate (%) is higher than when the mean squared error (MSE) is used. The difference between the mean squared error and other loss functions is that the loss value is large when the error is large in the mean squared error. From the results shown in Tables 6, 7, and 8, it was confirmed that it is preferable to select a loss function whose loss value is always equal to or less than the absolute value of the error, from the viewpoint of improving prediction accuracy.

[0119] (Influence of learning method) The effect of the learning method when constructing a model for predicting quality information of fresh concrete was verified. The accuracy rate (%) was compared between a case where prediction models M (prediction models M for each of the three qualities) were constructed individually for slump, slump flow, and air content by single-task learning, and a case where prediction model M that outputs predicted values ​​for the three qualities together was constructed by multi-task learning. In the verification, the above-mentioned 2765 training datasets and the above-mentioned 789 evaluation datasets were used. The verification results are shown in Tables 9, 10, and 11 below.

[0120] [Table 9]

[0121] [Table 10]

[0122] [Table 11]

[0123] In the multi-task learning column of Table 9, the accuracy rate (%) is calculated using the output value related to slump from prediction model M. In the multi-task learning column of Table 10, the accuracy rate (%) is calculated using the output value related to slump flow from prediction model M, and in the multi-task learning column of Table 11, the accuracy rate (%) is calculated using the output value related to air volume from prediction model M. From the results shown in Tables 9, 10, and 11, it can be seen that there is no significant difference in the accuracy rate (%) regardless of whether the learning method, single-task learning or multi-task learning, is adopted.

[0124] [Summary of this disclosure] The above-described method for predicting the quality of fresh concrete includes an acquisition step of acquiring input information including first information relating to at least one of the power load value of a mixer (114, 204) that produces fresh concrete by mixing concrete materials or mixes the fresh concrete, vibrations caused by the falling or flowing down of fresh concrete, and an image of the fresh concrete, and second information relating to a target quality of the fresh concrete when it is used, and a prediction step of predicting the quality of the fresh concrete based on a prediction model (M) previously constructed by machine learning to output quality information indicating the quality of the fresh concrete in response to input of the input information, and the input information acquired in the acquisition step.

[0125] When ready-mixed concrete is manufactured (for example, when it is manufactured in a factory or the like in which the manufacturing system (1) is installed), a target quality (at the site) of the ready-mixed concrete to be manufactured at the time of use is specified, and ready-mixed concrete is manufactured so as to satisfy the target quality. It has been found that by making such a target quality one of the pieces of input information for the prediction model (M) constructed by machine learning, the accuracy of the quality prediction is improved, as described above. It is easy to obtain the target quality as information. Therefore, in the above method, when predicting quality using the prediction model (M), it is possible to improve the accuracy of the quality prediction while avoiding the complication of work associated with quality prediction.

[0126] In the above-described method for predicting the quality of ready-mixed concrete, the quality of ready-mixed concrete may include one or more qualities of slump, slump flow, and air content. In this case, it is possible to easily manage whether the slump, slump flow, or air content of the produced ready-mixed concrete can satisfy the target quality when the ready-mixed concrete is used.

[0127] In the above-described method for predicting the quality of ready-mixed concrete, the quality of ready-mixed concrete may include two or more qualities of slump, slump flow, and air content. The prediction model (M) may be constructed to output two or more qualities in response to input of input information. As described above, it has been found that there is no significant difference in prediction accuracy between single-task learning to construct a prediction model (M) that outputs only one quality and multi-task learning to construct a prediction model M that outputs two or more qualities. By constructing the prediction model (M) by multi-task learning as in the above-described method, it is possible to simplify the work required to construct the model while maintaining prediction accuracy.

[0128] In the above-described method for predicting the quality of ready-mixed concrete, the prediction model (M) may be constructed by machine learning using a neural network that does not perform an operation for normalizing input information and an operation for missing some connections in the fully connected layer. In general, in machine learning using a neural network, it is considered preferable to perform an operation for normalizing input information and an operation for missing some connections in the fully connected layer from the viewpoint of stabilizing learning and suppressing overlearning. However, as described above, the presence or absence of these operations does not significantly affect the prediction accuracy, and it has been shown that for a specific quality (air content), the prediction accuracy tends to be higher when these operations are not performed. Therefore, by constructing the prediction model (M) in the above-described manner, it is possible to simplify the work required for constructing the model while maintaining the prediction accuracy.

[0129] In the above-described method for predicting the quality of ready-mixed concrete, the second information may include information indicating a target quality of the ready-mixed concrete at the time of shipment, in addition to information indicating a target quality of the ready-mixed concrete at the time of use. In a factory or the like in which the manufacturing system (1) is installed, the target quality at the time of shipment is set so as to satisfy the target quality of the ready-mixed concrete at the time of use (i.e., the required quality at the time of acceptance). Therefore, the target quality at the time of shipment correlates with the target quality of the ready-mixed concrete at the time of use. In the above-described method, the target quality at the time of shipment is further included in the input information of the prediction model (M), thereby making it possible to further improve the accuracy of the quality prediction.

[0130] In the above-described method for predicting the quality of ready-mixed concrete, the prediction model (M) may be constructed by machine learning using a neural network. In the machine learning using a neural network when constructing the prediction model (M), one selected from the group consisting of Adam, AdamBelief, Adamax, AdaBound, Adagrad, AMSGRAD, AMSBound, RMSprop, SgdW, Momentum, and Nesterov may be used as the weight update formula. As described above, it has been found that by using these update formulas, the prediction accuracy is higher than when other update formulas are used. Therefore, the above method can further improve the prediction accuracy.

[0131] In the above-described method for predicting the quality of ready-mixed concrete, the prediction model (M) may be constructed by machine learning using a neural network. In the machine learning using a neural network for constructing the prediction model (M), a function may be used as the loss function such that the loss value L(a) is always equal to or less than the absolute value of the error a. As described above, it has been found that the use of such a loss function improves prediction accuracy compared to the use of other loss functions. Therefore, the above method can further improve prediction accuracy.

[0132] The quality prediction program described above is a program for making a computer execute the above quality prediction method. Since the quality prediction program can make a computer execute the above quality prediction method, when predicting quality using a prediction model (M), it is possible to improve the accuracy of quality prediction while avoiding the complication of work involved in quality prediction.

[0133] The quality prediction device (10) described above includes an input information acquisition unit (24, 26) that acquires input information including first information related to at least one of a power load value of a mixer (114, 204) that manufactures ready-mixed concrete by mixing concrete materials or mixes ready-mixed concrete, vibration caused by the falling or flowing down of ready-mixed concrete, and an image of ready-mixed concrete, and second information related to a target quality of the ready-mixed concrete when it is used, a prediction model (M) that is constructed in advance by machine learning so as to output quality information indicating the quality of the ready-mixed concrete in response to input of the input information, and a prediction calculation unit (28) that predicts the quality of the ready-mixed concrete based on the input information acquired by the input information acquisition unit (24, 26). In the quality prediction device (10), when predicting quality using the prediction model (M), as in the above-mentioned quality prediction method, it is possible to improve the accuracy of the quality prediction while avoiding the complication of work associated with the quality prediction. [Explanation of symbols]

[0134] 1... manufacturing system, 10... control device, 24... operation information acquisition unit, 26... target quality information acquisition unit, 28... prediction calculation unit, 30... model construction unit, M... prediction model, 100... manufacturing apparatus, 114... mixer, 114a... stirring member, 114b... mixer drive unit, 204... mixer.

Claims

1. an acquisition step of acquiring input information including first information related to at least one of a power load value of a mixer that produces ready-mixed concrete by mixing concrete materials or mixes the ready-mixed concrete, vibration caused by the falling or flowing down of the ready-mixed concrete, and an image of the ready-mixed concrete, and second information related to a target quality of the ready-mixed concrete when it is used; A prediction process for predicting the quality of the ready-mixed concrete at the time of shipment based on a prediction model previously constructed by machine learning so as to output quality information indicating the quality of the ready-mixed concrete at the time of shipment in response to input of the input information and the input information acquired in the acquisition process; A method for predicting the quality of ready-mix concrete, including:

2. The quality of the ready-mix concrete at the time of shipment includes one or more qualities of slump, slump flow, and air content. The method for predicting quality of ready-mix concrete according to claim 1.

3. The quality of ready-mix concrete at the time of shipment includes two or more qualities of slump, slump flow, and air content, The prediction model is constructed to output the two or more qualities in response to the input information. The method for predicting quality of ready-mix concrete according to claim 2.

4. The prediction model is constructed by machine learning using a neural network that does not perform an operation to normalize the input information and an operation to cause some connections in a fully connected layer to be missing. The method for predicting quality of ready-mix concrete according to claim 1.

5. The second information includes information indicating a target quality at the time of shipment of the ready-mixed concrete in addition to information indicating a target quality at the time of use of the ready-mixed concrete. The ready-mixed concrete quality prediction method according to any one of claims 1 to 4.

6. The prediction model is constructed by machine learning using a neural network, In the machine learning using a neural network for constructing the prediction model, one selected from the group consisting of Adam, AdamBelief, Adamax, AdaBound, Adagrad, AMSGRAD, AMSBound, RMSprop, SgdW, Momentum, and Nesterov is used as a weight update formula. The method for predicting quality of ready-mixed concrete according to any one of claims 1 to 4.

7. The prediction model is constructed by machine learning using a neural network, In the machine learning using a neural network for constructing the prediction model, a loss function is used in which the loss value L(a) is always equal to or less than the absolute value of the error a. The method for predicting quality of ready-mixed concrete according to any one of claims 1 to 4.

8. A quality prediction program that causes a computer to execute the quality prediction method of ready-mixed concrete according to any one of claims 1 to 4.

9. A manufacturing process for producing ready-mix concrete; A quality prediction process for predicting the quality of the ready-mixed concrete at the time of shipment produced in the production process by the ready-mixed concrete quality prediction method according to any one of claims 1 to 4. A method for producing ready-mix concrete.

10. an input information acquisition unit that acquires input information including first information related to at least one of a power load value of a mixer that produces ready-mixed concrete by mixing concrete materials or mixes the ready-mixed concrete, vibration caused by the falling or flowing down of the ready-mixed concrete, and an image of the ready-mixed concrete, and second information related to a target quality of the ready-mixed concrete when it is used; A prediction calculation unit that predicts the quality of the ready-mixed concrete at the time of shipment based on a prediction model that is constructed in advance by machine learning so as to output quality information indicating the quality of the ready-mixed concrete at the time of shipment in response to input of the input information, and the input information acquired by the input information acquisition unit; A ready-mix concrete quality prediction device comprising:

Citation Information

Patent Citations

  • Method of predicting quality of ready mixed concrete

    JP2020144099A

  • Method for predicting quality of ready-mixed concrete

    JP2021124304A

  • Quality predication system for ready mixed concrete and quality prediction method for ready mixed concrete

    JP2023031937A

  • Ready mixed concrete quality prediction method, ready mixed concrete manufacturing method, and ready mixed concrete manufacturing system

    JP2023103030A

  • JPP7318093B

Cited By

  • Concrete slump control system and method combining AI monitoring and ERP

    CN120560023A

  • Quality prediction method for ready-mixed concrete, quality prediction program, and quality prediction device for ready-mixed concrete

    JP2025155760A