Road roller intelligent control method based on real-time state of pavement compaction quality

Abstract

The invention discloses a road roller intelligent control method based on a road surface compaction quality real-time state, an upper computer mainly judges compaction states and quality such as compaction degree, strength and modulus in a road surface compaction process in different rolling stages in situ through an in-situ detection vehicle, and a lower computer mainly aims at parameters fed back by the in-situ detection vehicle. According to the method, road surface compaction states and rolling process models in different rolling stages are learned and predicted, target function optimization considering the compaction degree, strength, modulus and the like is obtained, and the vibration frequency, amplitude, running speed, rolling times and the like of the road roller are adjusted in time, so that accurate regulation and control of the road surface compaction quality in different rolling stages are achieved. On the basis of comprehensively detecting the compaction quality of the asphalt pavement, automatic optimization of a rolling process considering target functions such as the compaction degree, the strength and the modulus can be completed, and the method has the advantages that pressing, detection and adjustment are carried out simultaneously, and the method is accurate and reliable.

Description

Technical Field

[0001] The present invention belongs to the technical field of automation and intelligent control of highway construction machinery, and in particular relates to an intelligent control method for a road roller based on the real-time status of road compaction quality. Background Art

[0002] At present, a lot of research has been done at home and abroad on the evaluation and control of the compaction quality status of asphalt pavements. For example, by drilling asphalt mixture core samples on site (CN202211590863.X), the density of the core samples is measured and compared with the standard density to evaluate the compaction of the asphalt pavement; the compaction of the asphalt pavement is evaluated by ground penetrating radar (CN202210285919.4); the mechanical strength of the asphalt pavement is evaluated by a compaction shear property tester (CN201910884633.6), etc. The intelligent compaction quality tester (ICCC) is used in combination with the acceleration sensor to intelligently control the vibration frequency and rolling speed of the roller in real time based on the compaction degree (CN202010955640.3), optimize the rolling process, etc.

[0003] However, there are the following prominent bottlenecks: 1) The evaluation index and design index of asphalt pavement are out of line: Currently, the evaluation index of intelligent compaction quality of asphalt pavement is compaction degree, while the evaluation index of theoretical design of asphalt pavement material structure is stiffness, strength, etc. Therefore, the evaluation and design indicators are not unified; 2) Single index control: The existing asphalt pavement compaction quality only focuses on compaction degree, and the index is superficial; in theory, good compaction quality requires not only high density of asphalt pavement, but also a reasonable skeleton structure and excellent mechanical properties, which are dense and represent strength. 3) Single stage control: The existing asphalt pavement compaction quality focuses mostly on the final compaction stage of asphalt pavement; there is a lack of guidance on the construction process of the initial compaction, re-compaction and final compaction stages of asphalt pavement, and there is still a lack of research on the construction process of pavement compaction based on dual indicators of volume and mechanics, which greatly limits the improvement of asphalt pavement compaction quality. Summary of the Invention

[0004] In response to the problems existing in the existing asphalt pavement compaction process control, the purpose of the present invention is to propose an intelligent control method for a roller based on the real-time status of the pavement compaction quality. The method can realize the adjustment of the compaction parameters (vibration frequency, amplitude, driving speed, direction, number of rolling passes, etc.) in different rolling stages through real-time detection of the pavement compaction status at different rolling stages and dynamic automatic rolling plan decision-making, and is particularly suitable for the construction process adjustment of the roller for newly built asphalt pavement.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions: A method for intelligently controlling a roller based on the real-time status of pavement compaction quality comprises: an upper computer inputs the compaction status of an asphalt pavement at different rolling stages to a lower computer through a human-machine interface / command operation and real-time detection by an in-situ detection vehicle; the lower computer performs model learning and prediction based on feedback parameters; and through objective function optimization, timely adjusts the roller's vibration frequency, amplitude, driving speed, and number of rolling passes for the asphalt pavement at different rolling stages, thereby achieving precise control of the pavement compaction quality at different rolling stages.

[0006] Specifically, the in-situ inspection vehicle detects the entire rolling stage, including the initial compaction, repeated compaction and final compaction stages, and can detect and output in real time the volume index compaction degree, mechanical index strength and modulus of the asphalt pavement status at different rolling stages.

[0007] Specifically, the lower computer includes a state fusion and prediction subsystem, a state optimization control subsystem, a roller travel subsystem and a roller compaction subsystem.

[0008] Specifically, the state fusion and prediction subsystem generates a compaction quality distribution heat map based on the data from the compaction quality real-time detection module, and then establishes an asphalt pavement compaction, strength and modulus evolution model that takes into account the roller's vibration frequency, amplitude, driving speed and number of rolling passes.

[0009] Specifically, the state optimization control subsystem constructs a multi-objective optimization function of compaction, strength and modulus, conducts deep learning of the LSTM prediction network based on different compaction stages, and automatically generates supplementary compaction processes.

[0010] Specifically, the roller travel subsystem uses V2X communication technology to transmit the evaluation results of the on-site inspection vehicle to the vehicle terminal (OBU) installed on the roller in real time, and receives the latitude and longitude of the on-site inspection vehicle.

[0011] Specifically, the roller compaction subsystem transmits the automatically generated supplementary compaction process motion instructions to the vehicle control system installed on the compaction equipment through the CAN protocol. After receiving the feedback results, the vehicle control system carries out optimized compaction control.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention uses a temperature sensor, a non-nuclear density meter, a lateral torsion test plate and a vertical loading test plate to detect and evaluate the temperature, compaction, strength and modulus of the asphalt pavement in real time.

[0013] 2. The present invention adjusts the rolling schemes at different rolling stages in real time through the multi-objective optimization function of compaction, strength and modulus, and ensures that the compaction, strength and modulus requirements are met in a refined manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is the architecture diagram of the intelligent control system of the road roller.

[0015] Figure 2 It is the intelligent control flow chart of the road roller. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] Take the re-pressing construction of a new asphalt pavement as an example: A method for intelligently controlling a roller based on the real-time status of pavement compaction quality comprises: an upper computer inputs the compaction status of an asphalt pavement at different rolling stages to a lower computer through a human-machine interface / command operation and real-time detection by an in-situ detection vehicle; the lower computer performs model learning and prediction based on feedback parameters; and through objective function optimization, timely adjusts the roller's vibration frequency, amplitude, driving speed, and number of rolling passes for the asphalt pavement at different rolling stages, thereby achieving precise control of the pavement compaction quality at different rolling stages. like Figure 1 As shown, the intelligent control system architecture of the road roller of the present invention consists of a host computer and a slave computer. When the host computer human-machine interface / command is issued, the in-situ inspection vehicle begins to detect the pavement compaction status during the re-compacting construction of the newly built asphalt pavement and provides information on the pavement compaction quality during the re-compacting phase. The slave computer then integrates and predicts the compaction status during the re-compacting phase, optimizes the compaction status, and plans the roller's travel and compaction plan.

[0018] The in-situ inspection vehicle covers the entire rolling phase, including initial, secondary, and final compaction. It can monitor and output in real time the volumetric compaction index, as well as the mechanical indicators of strength and modulus, for the asphalt pavement at each stage of compaction. The host computer human-machine interface / commands can be a mobile phone, industrial computer, or tablet.

[0019] The in-situ inspection vehicle measures asphalt pavement compaction status, including volumetric indicators such as compaction and mechanical indicators such as strength, modulus, and temperature. A single sampling time for these indicators does not exceed one minute; gridded pavement compaction, strength, and modulus are updated every 10-60 seconds with a resolution of no less than 0.1m x 0.1m. The in-situ inspection vehicle includes an asphalt pavement compaction quality assessment system, which uses real-time data collected to assess the compaction quality of the asphalt pavement. The in-situ inspection vehicle is equipped with an onboard unit (OBU), which transmits compaction status and quality information to the roller at the lower location via CAN protocol.

[0020] like Figure 1As shown in the figure, the lower computer includes the state fusion and prediction subsystem, the state optimization control subsystem, the roller travel subsystem and the roller compaction subsystem. The state fusion and prediction subsystem generates a compaction quality distribution heat map based on data from the real-time compaction quality detection module. It then develops an evolution model for the asphalt pavement's compaction, strength, and modulus, taking into account factors such as the roller's vibration frequency, amplitude, travel speed, and number of passes. The state optimization control subsystem constructs a multi-objective optimization function for compaction, strength, and modulus, and uses deep learning within an LSTM prediction network based on different compaction stages to automatically generate supplementary compaction processes. The roller travel subsystem, leveraging V2X communication technology, transmits the evaluation results of the in-situ inspection vehicle to the vehicle terminal (OBU) mounted on the roller in real time and receives the latitude and longitude coordinates of the in-situ inspection vehicle. The roller compaction subsystem transmits motion commands based on the automatically generated supplementary compaction process to the vehicle control system mounted on the compactor via the CAN protocol. Upon receiving this feedback, the vehicle control system implements optimized compaction control.

[0021] like Figure 2 The figure shows the intelligent control process for the roller of the present invention. After the in-situ inspection vehicle provides the compaction status of the recompaction inspection area, the inspection vehicle determines whether the recompaction volume and mechanical indicators meet the requirements. If so, the roller continues to move forward. If not, the roller vehicle terminal (OBU) receives the longitude and latitude coordinates of the in-situ inspection vehicle and moves to the location. Simultaneously, the roller performs deep learning of the LSTM prediction network based on the recompaction phase and optimizes the multi-objective optimization function for compaction degree, strength, and modulus to automatically generate a recompaction and supplementary rolling process. Furthermore, motion instructions based on the automatically generated supplementary rolling process are transmitted to the vehicle control system on the compacting tool via the CAN protocol. After receiving the feedback, the vehicle control system implements the optimized recompaction and supplementary rolling process.

[0022] Preferably, the roller is provided with a vehicle terminal (OBU), which can promptly receive feedback from the on-site inspection vehicle on the compaction status and quality whether it meets the standards through V2X communication technology. When it is detected that the compaction quality of a local area drops sharply by more than 5%, the emergency re-rolling mode is triggered.

[0023] Preferably, the LSTM prediction network deep learning can use the construction data of different rolling stages (≥500 groups of samples) to carry out deep learning of compaction quality (compaction degree, strength and modulus) based on ResNet-18, and adapt to the characteristics of different rollers and pavement materials (such as asphalt and aggregate type differences).

[0024] Preferably, the roller compaction process at different compaction stages includes the roller's vibration frequency, amplitude, travel speed, and number of compaction passes. Model predictive control optimization is performed every 30 seconds during the compaction process. Ultimately, through deep learning, the optimal compressive vibration frequency is determined to be 34 Hz, the amplitude is 1.88 mm, the excitation force is 280 kN, the travel speed is 1.8 km / h, and the number of compaction passes is 2.

Claims

1. An intelligent control method for a road roller based on the real-time status of road compaction quality, characterized in that: include: The upper computer inputs the compaction status of the asphalt pavement at different rolling stages to the lower computer through the human-machine interface / command operation and real-time detection of the in-situ detection vehicle. The lower computer combines the feedback parameters for model learning and prediction. Through objective function optimization, it timely adjusts the roller vibration frequency, amplitude, driving speed and number of rolling passes for the asphalt pavement at different rolling stages to achieve precise control of the pavement compaction quality at different rolling stages.

2. The intelligent control method for a road roller based on the real-time status of road compaction quality according to claim 1 is characterized in that: The in-situ inspection vehicle detects the entire rolling stage, including the initial compaction, repeated compaction and final compaction stages, and can detect and output in real time the volume index compaction degree, mechanical index strength and modulus of the asphalt pavement status at different rolling stages.

3. The intelligent control method for a road roller based on the real-time status of road compaction quality according to claim 2, characterized in that: The lower computer includes a state fusion and prediction subsystem, a state optimization control subsystem, a roller travel subsystem and a roller compaction subsystem.

4. The intelligent control method for a road roller based on the real-time status of road compaction quality according to claim 3 is characterized in that: The state fusion and prediction subsystem generates a compaction quality distribution thermogram based on data from the compaction quality real-time detection module, and then establishes an asphalt pavement compaction, strength, and modulus evolution model that takes into account the roller's vibration frequency, amplitude, driving speed, and number of rolling passes.

5. The intelligent control method for a road roller based on the real-time status of road compaction quality according to claim 4 is characterized in that: The state optimization control subsystem constructs a multi-objective optimization function of compaction, strength and modulus, conducts deep learning of the LSTM prediction network based on different compaction stages, and automatically generates supplementary compaction processes.

6. The intelligent control method for a road roller based on the real-time status of road compaction quality according to claim 5 is characterized in that: The roller travel subsystem is based on V2X communication technology, which transmits the evaluation results of the on-site inspection vehicle to the vehicle terminal (OBU) installed on the roller in real time, and receives the longitude and latitude of the on-site inspection vehicle.

7. The intelligent control method for a road roller based on the real-time status of road compaction quality according to claim 6, characterized in that: The roller compaction subsystem transmits the motion instructions based on the automatically generated supplementary compaction process to the vehicle control system installed on the compaction equipment through the CAN protocol. The vehicle control system carries out optimized compaction control after receiving the feedback results.

Images