Thread Content
Analysis of the Computer Simulation of the Automatic Leveling System in Asphalt Pavers Authors: Wang Xiaofeng, Chen Yong Abstract: Computer simulations are used to examine the operation of the automatic leveling control system in asphalt pavers, to optimize the adjustment methods of the controllers, and to analyze the impact of changes in the positions of sensors within the leveling system on its performance. This helps to improve the smoothness of asphalt pavements in practical applications. Keywords: PID control ； Asphalt paver ； Composite closed-loop control system ； Automatic leveling system: In the process of controlling road surface flatness, the automatic leveling control system of asphalt pavers plays a crucial role. The automatic leveling system of the asphalt paver uses a composite control system. A simple open-loop control system has advantages such as simple structure, low cost, and stable operation; it can achieve good results when the error is predictable. However, this control method cannot automatically correct the deviation of the controlled variable, and unknown disturbances have a significant impact on control accuracy. A simple closed-loop control system has the ability to automatically correct deviations in the controlled variable, resulting in high control accuracy. However, due to the presence of feedback, issues such as oscillations in the controlled variable or insufficient regulation can occur when the system parameters are not set properly. A composite control system adds a feedforward path (an open-loop system) on top of closed-loop control. Introducing a feedforward path not only has little impact on the performance of the closed-loop system, but can also **improve the control accuracy of the system. Before the effects of surface irregularities reach the bottom surface of the smoothing plate, they are transmitted from the aircraft body to the error detector, and the system begins to perform pre-adjustment. When the smoothing plate moves into this error zone, the relative error has already been reduced due to this pre-adjustment; as a result, the error transmitted back to the detector from the smoothing plate becomes very small, which in turn reduces the range required for further adjustment. Using computer technology to simulate its actual operating conditions is of great significance both from a research perspective and for guiding actual paving work. 1 Establishment of the system mathematical model: Based on existing research findings, by ignoring the disturbances at the traction point and the screed, the longitudinal leveling system of the asphalt paver can be simplified as shown in Figure 1. The regulator represents an automatic level control device, whose function is to adjust and amplify signals. In the research on digital controllers, an appropriate combination of proportional (P), integral (I), and differential (D) control is used as a substitute. A sensor can be considered as a proportional element given by G1(s)=k, where k is the sensor’s proportional constant. The actuator consists of a servo valve and a double-acting hydraulic cylinder. The transfer function of the proportional servo valve can be expressed as a first-order element, where kq represents the flow gain of the servo valve ； τ1 is the servo valve time constant. The transfer function of the hydraulic cylinder is given by the formula, where ωv is the hydraulic natural frequency of the valve-controlled cylinder ； Mt is the mass of the ironing plate supported by a single cylinder ； V is the volume of the equivalent hydraulic cylinder ； β is the equivalent volume modulus of elasticity ； ξV is the hydraulic damping ratio of the valve-controlled cylinder ； #It is the effective area of the hydraulic cylinder. The ironing plate can be modeled as a first-order inertial element, resulting in the following transfer function; here, τ2 is the time constant of the ironing plate, with τ2 = L/v ； L is the effective length of the traction arm ； v is the paving speed. The two feedback loops are given by the formula, where c is the distance from the sensor position to the traction point. 2 Computer Simulation and Result Analysis: System parameters are determined based on existing models; the fuel supply pressure is 17 Mpa ； The diameter of the cylinder piston is D=100mm, the diameter of the piston rod is d=50mm, and the piston stroke is L=800mm ； The elastic modulus of the oil, β, is 700 MPa ； The weight of the ironing plate is 16,000 kg ； Other relevant parameters are determined based on the actual situation. First, draw the Bode plot of the control system, as shown in Figure 2. The true signal that constitutes road surface irregularity is a white noise signal; however, depending on the paving conditions of the paver and the filtering effect of the screed, only the error information related to the road surface in the form of low-frequency signals has an impact on the system, while high-frequency error signals are filtered out by the screed. The determination of the frequency of the road surface error signal depends on two factors: the paving speed and the width of the screed. Based on the recommended paving speed and screed width for current pavers, only road surface error signals below 0.5 Hz are considered valid signals. Based on the frequency characteristics shown in the system Bode plot, it can be seen that the system generally matches the actual auto-leveling operating conditions, indicating that the constructed simulation system model is effective. 2.1 Optimization analysis of the regulator control method: This paper focuses on optimizations based on the traditional PID control approach. The traditional PID algorithm is suitable for control systems in general industrial processes, offering higher control strength and precision compared to rule-based algorithms. The three elements of proportional control P, integral control I, and derivative control D can be combined to form three types of controllers: PD (proportional derivative), PI (proportional integral), and PID (proportional integral derivative). Given that the ironing plate mechanism is a relatively large first-order inertial element with significant lag, lag correction is necessary. In Figure 3, curve a represents the step response curve of the system using P control, while curve b represents the step response curve of the system using PD control ； Figure 4 shows the response curves of the system to sine signals of the same frequency under the two control methods. Curve c represents the response curve of the system using P control for sine signals, while curve d represents that of the system using PD control. From curve a in Figure 3, it can be seen that a simple proportional control system results in large overshoot and a long settling time, whereas curve b shows that the step response after adopting PD control corresponds to the ideal condition desired for the system. Figure 4 shows that the time-domain response of the system to the same sine signal varies significantly under different tuning methods. The d-curve indicates that with proportional-differential (PD) control, the system has good stability and low output lag. 2.2 Influence of sensor position on the system: The deviation detected by the sensor is caused by the vertical movement of the traction point and the ironing plate. The curve shown in Figure 5 depicts how the system’s step response changes as the C value (the distance between the sensor and the traction point) varies. As the C value increases, the overshoot grows, making it difficult for the system to stabilize. In actual operating conditions, when the sensor is located further back, it is possible to accurately detect the force that has just been applied to the layer, but it is not possible to predict the effects of subsequent external disturbances. Additionally, due to the lagging effect of the system, it tends to oscillate, resulting in poor stability. However, if the sensor is placed near the traction point, it can provide maximum sensitivity in the vertical direction of the traction point, but it cannot detect the vertical movement of the paver caused by factors such as paving speed, mixture density, and mixture temperature. The curve shows that when the sensor is positioned between the traction point and the ironing plate, the system response tends to be ideal. Based on the simulation and experimental results, a sensor is generally placed in front of the ironing plate at a position roughly equivalent to 1/3 of the length of the boom. At this point, the system’s performance will be fully utilized. 3 Conclusion Based on the mathematical model of the automatic leveling control system for pavers, computer-based optimization of the control methods for such systems is feasible, which holds certain value for the development of digital leveling controllers. The change in the position of sensors in the paver’s automatic leveling system has a crucial impact on the system’s performance; therefore, the analysis presented in this paper holds significant practical value for the engineering application of pavers, as it helps to improve the smoothness of road surfaces.