Online updating method and system for concrete slump model of concrete mixer

Abstract

The invention relates to an online updating method and system for a concrete slump model of a concrete mixer, and belongs to the technical field of engineering equipment. During concrete transportation, after concrete loading and before concrete unloading, at least the torque of a loading driving motor of a concrete mixer and the rotating speed of a tank body are collected when the tank body rotates, and meanwhile, the manually measured artificial concrete slump of the same concrete sample is obtained; updating and training the slump detection model according to the acquired torque of the upper driving motor and the rotating speed of the tank body; the slump detection model is obtained by training based on a machine learning model. The sample library is continuously updated in the transportation of the concrete mixer, so that the concrete slump detection accuracy of the slump detection model is improved.

Description

Technical Field

[0001] The present invention relates to an online update method and system for a concrete slump model of a concrete mixer truck, and belongs to the technical field of engineering equipment. Background Art

[0002] In civil engineering construction, fresh concrete is required to have good workability. For example, it is necessary to keep the fresh concrete from stratifying, segregating, bleeding, etc. Concrete structures with good slump, density, and uniformity are convenient for construction operations, including mixing, transportation, pouring, vibrating, and pumping.

[0003] Currently, the transportation distance and time of concrete mixer trucks are relatively long. Due to various reasons such as chemical reactions, water evaporation, and aggregate water absorption of concrete clinker, the free water content decreases, resulting in slump loss over time of the concrete. Even if the workability of the ready-mixed concrete is tested at the time of leaving the factory, due to uncontrollable factors during transportation, the workability of the concrete may change unpredictably, and the workability of the concrete at the time of delivery is very likely to deviate from the expected control range and cannot meet the construction requirements. Therefore, the transportation process of ready-mixed concrete is particularly crucial for the delivery quality of concrete, and the change in slump directly affects the construction quality of buildings. It is not easy to detect the slump during the transportation of the mixer truck to the construction site. Once construction quality problems occur, it is impossible to accurately determine whether the responsibility lies with the concrete supplier or the construction party.

[0004] Chinese patent document with publication number CN119006444A discloses a concrete slump prediction method and device based on image and flow velocity estimation. This method creates a set of stirring image sequences in the mixer, and through machine learning training based on image information and flow velocity, a fully connected neural network for judging the weight of slump values is obtained to achieve real-time detection of concrete slump. However, this method not only requires setting up a camera but also ensuring a light source, with high costs and difficult operation and maintenance.

[0005] Chinese patent document with publication number CN113899887A discloses a method and system for testing the time-varying workability of concrete based on driving torque. This method establishes a sample library of the difference between the maximum and minimum values of the driving shaft torque data of the concrete tank truck body for the workability of fresh concrete; based on the sample library, the workability of the fresh concrete is obtained, and the workability of the fresh concrete is displayed to complete the time-varying test of the workability of concrete based on driving torque. However, this method does not mention the maintenance and update training of the sample library, and there is a problem that the accuracy of detecting the slump of concrete is not high due to the singularity of the sample library in a complex environment. Summary of the Invention

[0006] The object of the present invention is to provide an online update method and system for the concrete slump model of a concrete mixer truck, so as to solve the problem of low accuracy in detecting the concrete slump due to the singularity of the sample library in a complex environment.

[0007] To achieve the above object, the solution of the present invention includes: An online update method for the concrete slump model of a concrete mixer truck according to the present invention. During concrete transportation, when the concrete is loaded into the truck and the tank is rotating, at least the torque of the upper-mounted driving motor and the tank rotation speed of the concrete mixer truck are collected. At the same time, the manually measured artificial concrete slump of the same concrete sample is obtained, and the slump detection model is updated and trained by using the artificial concrete slump, the collected torque of the upper-mounted driving motor and the tank rotation speed; the slump detection model is trained based on a machine learning model.

[0008] Further, the tank temperature is also collected when the concrete is loaded into the truck and the tank is rotating; the slump detection model is updated and trained by using the artificial concrete slump, the collected torque of the upper-mounted driving motor, the tank rotation speed and the tank temperature.

[0009] Further, when obtaining the artificial concrete slump, the concrete grade and the volume are also obtained at the same time; the slump detection model is updated and trained by using the artificial concrete slump, the simultaneously obtained concrete grade and volume, and the collected torque of the upper-mounted driving motor, the tank rotation speed and the tank temperature.

[0010] Further, the artificial concrete slump, the concrete grade and the volume are input through a control operation display on the vehicle, and the artificial concrete slump, the concrete grade and the volume, and the collected torque of the upper-mounted driving motor, the tank rotation speed and the tank temperature are uploaded to the vehicle networking platform through a vehicle-mounted communication device for updating and training the slump detection model.

[0011] The present invention also provides an online update method for the concrete slump model of a concrete mixer truck. During concrete transportation, when the concrete is unloaded from the truck and the tank is rotating, at least the torque of the upper-mounted driving motor and the tank rotation speed of the concrete mixer truck are collected. At the same time, the manually measured artificial concrete slump of the same concrete sample is obtained, and the slump detection model is updated and trained by using the artificial concrete slump, the torque of the upper-mounted driving motor and the tank rotation speed.

[0012] Further, the tank temperature is also collected when the concrete is unloaded from the truck and the tank is rotating; the slump detection model is updated and trained by using the artificial concrete slump, the collected torque of the upper-mounted driving motor, the tank rotation speed and the tank temperature.

[0013] Further, when obtaining the slump of the artificial concrete, the concrete grade and the quantity are also obtained simultaneously; the slump detection model is updated and trained by using the slump of the artificial concrete, the simultaneously obtained concrete grade, quantity, and the torque of the upper-mounted driving motor, the tank rotation speed, and the tank temperature collected.

[0014] Further, the slump of the artificial concrete, the concrete grade, and the quantity are input through a control operation display on the vehicle, and the slump of the artificial concrete, the concrete grade, and the quantity, as well as the torque of the upper-mounted driving motor, the tank rotation speed, and the tank temperature collected are uploaded to the vehicle networking platform through a vehicle-mounted communication device for updating and training the slump detection model.

[0015] Finally, the present invention also provides an online update system for the concrete slump model of a concrete mixer truck, including a processor. The online update system for the concrete slump model of the concrete mixer truck is used to execute a computer program to implement the steps of the online update method for the concrete slump model of a concrete mixer truck.

[0016] The beneficial effects of the present invention are as follows: The present invention provides an online update method and system for the concrete slump model of a concrete mixer truck. The torque of the upper-mounted driving motor and the tank rotation speed are collected through the automatic data collection system of the concrete mixer truck, and the slump of the artificial concrete is also obtained. The slump is calculated in real time through a pre-trained slump detection model. At the same time, the slump detection model is continuously updated and trained through the results of the artificial measurement of the slump before and after the concrete mixer truck is loaded and unloaded each time it transports concrete. The method of the present invention continuously updates the sample library during the transportation of the concrete mixer truck, improves the accuracy of the slump detection model for detecting the slump of the concrete, and further realizes more accurate real-time monitoring of the slump of the concrete during the transportation process, improves the quality of concrete delivery and customer satisfaction, and reduces the supply-demand contradiction. When the slump deviation detected by the slump detection model is too large, a warning is issued to timely remind the driver to add water reducing agent, etc., to ensure that the slump is within the normal range.

[0017] The online update system for the concrete slump model of the concrete mixer truck can achieve the same beneficial effects as the above-mentioned online update method for the concrete slump model of the concrete mixer truck. Description of the Drawings

[0018] Figure 1 is a partial schematic diagram of an electric upper-mounted system for online monitoring of the change of concrete slump; Figure 2 is a schematic flow chart of the slump detection method. Detailed Embodiments

[0019] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be described in detail below in conjunction with the accompanying drawings and embodiments.

[0020] The inventive concept lies in an online update method and system for the concrete slump model of a concrete mixer truck. During concrete transportation, after the concrete is loaded into the tank, at least the torque of the upper-mounted drive motor and the tank rotation speed of the concrete mixer truck are collected while the tank is rotating. At the same time, the manually measured artificial concrete slump of the same concrete sample is obtained. The slump detection model is updated and trained using the artificial concrete slump, the collected torque of the upper-mounted drive motor and the tank rotation speed. And before the concrete is unloaded, at least the torque of the upper-mounted drive motor and the tank rotation speed of the concrete mixer truck are collected while the tank is rotating. At the same time, the manually measured artificial concrete slump of the same concrete sample is obtained. The slump detection model is updated and trained using the artificial concrete slump, the torque of the upper-mounted drive motor and the tank rotation speed. The slump detection model is trained based on a machine learning model. The sample library is continuously updated during the transportation of the concrete mixer truck to improve the accuracy of detecting the concrete slump.

[0021] Method Embodiment 1: An online update method for the concrete slump model of a concrete mixer truck according to the present invention is applicable to a concrete mixer truck with an electric upper-mounted system that uses a pre-trained concrete slump model to predict the concrete slump online in real time during vehicle operation. The electric upper-mounted system can effectively monitor the change in the rotation torque of the mixing drum during concrete transportation. The concrete slump model reflects the fluidity of the concrete using the motor torque, thereby indirectly monitoring the change in the concrete slump.

[0022] The electric upper-mounted system is as Figure 1 shown, and includes an upper-mounted motor 11 connected to a speed reducer 12, a drive motor controller 14 connected to a drive motor assembly of the mixing drum 13 and an upper-mounted controller 15, an upper-mounted controller 15 for controlling the upper-mounted state and rotation speed, a communication device 16 (Tbox is used in this embodiment) for communicating with the upper-mounted controller, a vehicle networking platform for performing wireless data exchange with the communication device 16 and performing data processing and analysis, a control operation display 17 installed in the cab, a user terminal 18 (Web / APP), and a temperature sensor 19 installed on the mixing drum 13.

[0023] The concrete mixer truck includes an automatic data acquisition system and a manual data entry acquisition system. Among the parameters obtained by the automatic data acquisition system, there are vehicle operation time, location, vehicle speed, mixer motor speed (corresponding to the tank body speed), mixer motor torque, and tank body temperature information. The mixer motor speed and mixer motor torque are collected by the mixer controller 15. The parameters obtained by the manual data entry acquisition system include concrete grade, volume, manual slump test time and results. The data of the automatic data acquisition system and the manual data entry acquisition system are transmitted to the Tbox, and the Tbox transmits them to the vehicle networking platform. On the vehicle networking platform, the collected data is used to calculate the slump test results through the trained model. Specifically, the license plate number is bound to the sim card number of the Tbox. On the vehicle networking cloud platform, it is judged which vehicle the corresponding data belongs to according to the sim card number of the Tbox that uploads the data. After the corresponding concrete slump result is detected by the concrete slump detection model, the result is sent to the corresponding vehicle.

[0024] The motor and the motor controller of the electric mixer system of the above-mentioned mixer truck can be integrated together structurally; Preferably, there can be only one vehicle controller and electric mixer controller, or there can be both; Preferably, the Tbox can be an independent component or can be integrated with other components structurally; The concrete slump detection model mentioned in this method can perform real-time online measurement of slump. During the transportation of the concrete mixer truck, the vehicle control operation display 17 or the user terminal 18 of the vehicle driver can display the slump in real time. When the concrete slump deviation is large, the driver can be reminded to add water reducing agent and other operations. At present, the slump detection model of the concrete mixer truck trains and measures the planned sample library. The data source of the sample library is single. As the vehicle usage years increase and the environmental complexity increases, the accuracy of the real-time detection results of the concrete slump model gradually decreases.

[0025] In the concrete transportation in daily use of this method, after the concrete is loaded, at least the torque of the mixer drive motor and the tank body speed of the concrete mixer truck are collected, and at the same time, the manually measured artificial concrete slump of the same concrete sample is obtained. The artificial concrete slump, the torque of the mixer drive motor and the tank body speed are used to update and train the slump detection model; the slump detection model is obtained by training based on a machine learning model. In the daily transportation of the concrete mixer truck, the sample library is continuously updated, and the concrete slump detection model is continuously updated and trained to improve the accuracy of detecting the concrete slump.

[0026] The steps of the online update method of the concrete slump model of the concrete mixer truck in this embodiment are as follows: 1) Model update training: Before the concrete mixer truck is loaded and leaves the factory, the same concrete sample retained or extracted from the concrete mixer truck is manually measured to obtain the manual concrete slump. Alternatively, after measuring the slump, the concrete is loaded back into the concrete mixer truck; the manual concrete slump is input at the user end of the vehicle networking platform (such as the control operation display on the vehicle). 2) At the same time, after the concrete is loaded, the motor controller of the electric upper loading system of the concrete mixer truck automatically and real-time obtains the torque of the upper loading drive motor and the tank rotation speed of the concrete mixer truck. 3) Update and train the data in steps 1) and 2) in the concrete slump detection model.

[0027] As a specific implementation method, the manual concrete slump is input into the vehicle through the control operation display on the vehicle; the torque of the upper loading drive motor and the tank rotation speed of the concrete mixer truck collected by the upper loading controller are uploaded through the vehicle's Tbox. The torque of the upper loading drive motor, the tank rotation speed, and the manual concrete slump of the concrete mixer truck are obtained on the vehicle networking cloud platform to update and train the concrete slump detection model based on machine learning.

[0028] As other implementation methods, before the concrete mixer truck is loaded and leaves the factory, while obtaining the manually measured concrete slump, the concrete grade and volume are also obtained and input into the vehicle together through the control operation display on the vehicle; at the same time, the motor controller automatically and real-time obtains the torque of the upper loading drive motor and the tank rotation speed of the concrete mixer truck, and the tank temperature can also be further collected. The feature vectors of the concrete slump detection model are increased to improve the accuracy of detecting the concrete slump.

[0029] Specifically, relevant data is input into Table 1 at the user end of the vehicle networking platform, including time, tank temperature, license plate number, volume, concrete grade, manual detection time, manual detection slump value, motor torque, and tank rotation speed, as shown in the following table: Table 1 Date Time License Plate Number Volume Concrete Grade Motor Torque Tank Rotation Speed Tank Temperature Manual Inspection Time Manual Inspection Slump Value Automatically Calculated Slump Value The data in Table 1 is uploaded from the Tbox to the vehicle networking cloud platform to update and train the concrete slump detection model based on machine learning.

[0030] The concrete slump detection model mentioned in this method is updated, trained, and optimized for different concrete grades, volumes, slump detection values, and slump loss over time. When the result of the concrete slump detection model is consistent with the manual detection result, it can be considered that the model iteration is completed and fully used to replace the manual detection. When the slump deviation is too large and a warning is issued, the driver is reminded to add water reducing agent in time to ensure that the slump is within the normal range.

[0031] Method Embodiment 2: An online update method for the concrete slump model of a concrete mixer truck according to the present invention, based on the method embodiment 1, during the transportation of concrete in daily use, at least the torque of the upper-mounted drive motor and the tank rotation speed of the concrete mixer truck are collected before the concrete is unloaded, and at the same time, the manually measured artificial concrete slump of the same concrete sample is obtained, and the slump detection model is updated and trained by using the artificial concrete slump, the torque of the upper-mounted drive motor and the tank rotation speed; the slump detection model is obtained by training based on a machine learning model. The sample library is continuously updated during the daily transportation of the concrete mixer truck, and the concrete slump detection model is continuously updated and trained to improve the accuracy of detecting the concrete slump.

[0032] The steps of the online update method for the concrete slump model of the concrete mixer truck in this embodiment are as follows: 1) Model update training: Before the concrete is unloaded from the concrete mixer truck, the same concrete sample extracted from the concrete mixer truck is manually measured to obtain the artificial concrete slump, and is input at the user end of the vehicle networking platform (such as the control operation display on the vehicle). 2) At the same time, before the concrete is unloaded, the motor controller of the electric upper-mounted system of the concrete mixer truck automatically and real-time obtains the torque of the upper-mounted drive motor and the tank rotation speed of the concrete mixer truck. 3) Update and train the data in steps 1) and 2) in the concrete slump detection model.

[0033] As a specific implementation manner, the artificial concrete slump is input into the vehicle through the control operation display on the vehicle; the torque of the upper-mounted drive motor and the tank rotation speed of the concrete mixer truck collected by the upper-mounted controller; are uploaded through the Tbox of the vehicle, and the torque of the upper-mounted drive motor, the tank rotation speed and the artificial concrete slump of the concrete mixer truck are obtained on the vehicle networking cloud platform, and the concrete slump detection model based on machine learning is updated and trained.

[0034] As other implementation manners, before the concrete mixer truck finishes unloading, while obtaining the manually measured concrete slump, the concrete grade and the volume are also obtained, and are input into the vehicle through the control operation display on the vehicle together; at the same time, the motor controller automatically and real-time obtains the torque of the upper-mounted drive motor and the tank rotation speed of the concrete mixer truck, and the tank temperature can also be further collected. The feature vectors of the concrete slump detection model are increased to improve the accuracy of detecting the concrete slump.

[0035] Specifically, relevant data is input at the user end of the vehicle networking platform, including time, tank temperature, license plate number, volume, concrete grade, manual detection time, manual detection slump value, motor torque and tank rotation speed. The data is uploaded from the Tbox to the vehicle networking cloud platform, and the concrete slump detection model based on machine learning is updated and trained.

[0036] The concrete slump detection model mentioned in this method is updated, trained, and optimized according to different concrete grades, volumes, slump detection values, and slump loss over time. When the result of the concrete slump detection model is consistent with the manual detection result, it can be considered that the model iteration is completed and it can completely replace manual detection. When there is an early warning of excessive slump deviation, the driver is reminded in time to add water-reducing agent, etc., to ensure that the slump is within the normal range.

[0037] Method Embodiment 3: As an optimal implementation method, as Figure 2 shown, every time the concrete mixer truck transports, the manual detection results before and after loading are used to update and train the concrete slump detection model twice. The steps are as follows: Step 21, automatically and real-time obtain the vehicle information, torque of the upper-mounted drive motor, tank rotation speed, and tank temperature of the mixer truck through the motor controller, and manually input the concrete grade and volume. According to the slump detection model, the concrete slump detection result is output in real time; Step 22, model update training: Before the concrete mixer truck finishes loading and leaving the factory, manually measure the slump result for the first time and input it at the user end of the vehicle networking platform, including the concrete grade, volume, the time and result of the first manual slump detection. The data is used to perform the first update training on the detection model; Step 23, before the concrete mixer truck unloads, manually measure the slump result for the second time and input it at the user end of the vehicle networking platform, including the concrete grade, volume, the time and result of the first manual slump detection. The data is used to perform the second update training on the detection model; Step 24, the slump detection model is updated, trained, and optimized according to different concrete grades, volumes, slump detection values, and slump loss over time. When the result of the slump detection model is consistent with the manual detection result, it can be considered that the model iteration is completed and it can completely replace manual detection. When there is an early warning of excessive slump deviation, the driver is reminded in time to add water-reducing agent, etc., to ensure that the slump is within the normal range.

[0038] System Embodiment: This embodiment provides a technical solution for an online update system of a concrete slump model of a concrete mixer truck, including a processor. This system is used to execute a computer program to implement the steps of the above-mentioned online update method of the concrete slump model of the concrete mixer truck.

[0039] Since the specific implementation process and principle of the online update system of the concrete slump model of the concrete mixer truck in this embodiment have been described in detail in the method embodiment, no more details will be elaborated here.

Claims

1. A method for online updating of a concrete slump model of a concrete mixer truck, characterized in that: During concrete transportation, after the concrete is loaded onto the truck, at least the torque of the upper drive motor and the tank speed of the concrete mixer truck are collected when the tank rotates, and the manually measured slump of artificial concrete of the same concrete sample is obtained at the same time. The slump of artificial concrete and the collected torque of the upper drive motor and the tank speed are used to update the training of the slump detection model; the slump detection model is obtained based on the training of the machine learning model.

2. The method for online updating of the concrete slump model of a concrete mixer truck according to claim 1, characterized in that: After the concrete is loaded, the tank body temperature is also collected when the tank body rotates; the slump detection model is updated and trained using the artificial concrete slump, as well as the collected upper driving motor torque, tank body rotation speed and tank body temperature.

3. The online updating method of the concrete slump model of the concrete mixer truck according to claim 2, characterized in that: When obtaining the slump of the artificial concrete, the concrete grade and volume are also obtained at the same time; the slump detection model is updated and trained using the artificial concrete slump, the concrete grade and volume obtained at the same time, and the collected upper drive motor torque, tank speed and tank temperature.

4. The method for online updating of the concrete slump model of a concrete mixer truck according to claim 3, characterized in that: The slump of artificial concrete, concrete grade and volume are input through the control operation display on the vehicle, and the slump of artificial concrete, concrete grade and volume as well as the collected upper drive motor torque, tank speed and tank temperature are uploaded to the vehicle networking platform using the vehicle communication device to update and train the slump detection model.

5. A method for online updating of a concrete slump model of a concrete mixer truck, characterized in that: During concrete transportation, at least the upper drive motor torque and tank speed of the concrete mixer truck are collected when the tank rotates before the concrete is unloaded, and the manually measured artificial concrete slump of the same concrete sample is obtained. The slump detection model is updated and trained using the artificial concrete slump, the upper drive motor torque and the tank speed.

6. The method for online updating of the concrete slump model of a concrete mixer truck according to claim 5, characterized in that: Before the concrete is unloaded, the tank body temperature is also collected when the tank body rotates; the slump detection model is updated and trained using the artificial concrete slump, as well as the collected upper drive motor torque, tank body rotation speed and tank body temperature.

7. The method for online updating of the concrete slump model of a concrete mixer truck according to claim 6, characterized in that: When obtaining the slump of the artificial concrete, the concrete grade and volume are also obtained at the same time; the slump detection model is updated and trained using the artificial concrete slump, the concrete grade and volume obtained at the same time, and the collected upper drive motor torque, tank speed and tank temperature.

8. The method for online updating of the concrete slump model of a concrete mixer truck according to claim 7, characterized in that: The slump of artificial concrete, concrete grade and volume are input through the control operation display on the vehicle, and the slump of artificial concrete, concrete grade and volume as well as the collected upper drive motor torque, tank speed and tank temperature are uploaded to the vehicle networking platform using the vehicle communication device to update and train the slump detection model.

9. An online updating system for a concrete slump model of a concrete mixer truck, comprising a processor, characterized in that: The online updating system of the concrete slump model of the concrete mixer truck is used to execute a computer program to implement the steps of the online updating method of the concrete slump model of the concrete mixer truck according to any one of claims 1 to 8.

Images