Abstract: The design steps of the electro-hydraulic servo PID control system for a chamfering machine tool are described in detail. A comparative study of the performance of PID control, fuzzy control, and fuzzy control + PID control systems is conducted through simulation. Keywords : chamfering machine tool; electro-hydraulic servo system; PID control; fuzzy control. A steel plant's steel pipe chamfering and deburring process requires a high degree of automation, strong load-bearing capacity, and good system rigidity (no tool deflection under heavy load). After repeated comparisons, an electro-hydraulic servo system was selected as the machine tool's feed drive mechanism. 1 Calculation of Basic Hydraulic and Electrical Parameters This system is a typical valve-controlled cylinder system. Known operating conditions: 1) The weight of the hydraulic cylinder and tool holder under no-load conditions is 2 t, i.e., M[sub]1[/sub] = 2000 kg; 2) The cutting force is approximately 15 kN, and 20 kN is used in the simulation. Considering factors such as fluctuations in working conditions, the simulated load during the simulation is set as F=20000+50sin(50t) (N). 1.1 Determination of hydraulic parameters 1) The hydraulic cylinder is a T-type structure, and the stroke (rapid advance + coarse feed + fine feed) during operation is approximately 83mm. 1.2 Determination of electrical parameters 1) The displacement corresponding to each volt of voltage command is K=198 (V/m). 2) The rated current of the valve is 60 mA. Therefore, the output limit of the PID controller is ±60mA . 2 Transfer function of the electro-hydraulic servo system The block diagram of the electro-hydraulic servo system is obtained (see Figure 1). 4) The open-loop transfer function of the electro-hydraulic servo system is: From this, the BODE diagram can be obtained (see Figure 2), and it can be seen that due to the addition of the PID controller, the integral element of the original system's open-loop transfer function changes from first-order to second-order, thus improving the steady-state accuracy and making the system free from static error in response to position commands. [align=center]Figure 2 Open-loop system BODE diagram[/align] According to the definition of amplitude and phase angle stability margin, it can be seen from Figure 2 that the amplitude margin G[sub]m[/sub] = 18.4 dB and the phase angle margin P[sub]m[/sub] = 36.3[sup]o[/sup]. Therefore, the system has good stability margin and certain anti-interference ability and robustness. 3 Input instruction simulation Figure 3 shows the design block diagram of the instruction input system. Through the switching of 3 logic switches, the input instruction system is formed, that is, when t When t1 or t=t1, the input command is the coarse feed rate; when t>t2, the input command is the fine feed rate; when t>t3, the input command is zero, causing the system to fast back. 4 Simulation of PID Control 1) Step Response of Electro-hydraulic Servo System Figure 4 shows the step response of the system under different loads, where Lm represents the displacement of the worktable. Figure 4a shows the load value F=20 kN, Figure 4b shows the load value F=10 kN, and Figure 4c shows the load value suddenly increasing from 10 kN to 20 kN, and then to 30 kN. It can be seen that there is no overshoot under various loads. The transition time is slightly longer when the load is large, and the curve does not jitter or backtrack when the load changes, indicating that the system has sufficient stiffness. 2) Processing different pipe materials: Based on the cutting parameter table, two different pipe materials and several typical pipe diameters D were selected and simulated under a system load of 20 kN to obtain the response curves (see Figure 5). 5. Fuzzy control (Fc) of electro-hydraulic servo system : Given that the cutting force changes nonlinearly during processing, the load and cutting feed also change when processing different steel pipes, so the coefficients of PID control cannot always be in the optimal state. Fc is robust, does not require a mathematical model, and is simple and time-saving to control. The Fc algorithm was used for simulation research below. Figure 6 is a block diagram of FC-PID joint control. By switching, Fc can work independently, or it can automatically switch to PID control when the error e reaches a certain small value. The FC-optimized control obtained by a large number of PD simulations using the aforementioned method is shown in Table 1. In the table, NL, NM, and iNS represent negative large, negative medium, and negative small, ZO represents zero, and PS, PM, and PL represent positive small, positive medium, and positive large. Figure 7 shows the membership functions of the fuzzy control. Figures 7a, 7b, and 7c are the membership functions of the error e, the rate of change of error c, and the control signal n, respectively. The fuzzy set membership functions of the error e and the output control signal n are uniformly distributed on the universe of discourse, while de ~ dr: c are non-uniformly distributed. To improve control accuracy, a high resolution is used when the error is small. Figure 8 shows the step input simulation curve of the electro-hydraulic servo system obtained by the above fuzzy control. For comparison, Figure 8b shows the simulation results of PID control. It can be seen from Figure 8 that the fuzzy control has a small overshoot, and the output performance is slightly worse than that of the PID control. This is because the Fc interval is coarser, the operation is discontinuous, and the control accuracy is low, while PID control is continuous control and operates under optimized system conditions. 6 Fuzzy Control Plus PID Control As mentioned above, FC has its drawbacks, especially when the error e is already small, discontinuous control cannot reduce e to zero. Therefore, after Fc quickly reduces e to a very small value (adjustable), the system in Figure 6 automatically switches to PID, so that e is continuously controlled to zero, thereby improving dynamic quality and system accuracy. When the switching value (threshold value) is set to 0.5, the control result is shown in Figure 8c. Comparing with Figure 8b, it can be seen that Fc+PID control is roughly the same as PID control alone. However, even so, when the system parameters and external load change significantly, PID control cannot work under the original optimized state, while FC+PID control, due to the strong robustness of FC, can still obtain good control results. 7 Conclusions 1) Through simulation analysis, the system is stable and reliable, and the stability reserve meets the requirements; 2) The PID optimization parameters are: proportional Kp = 0.46, integral Ki = 0.006, derivative Kd = 0.0085. When the PID controller selects the above parameters, it reaches the critical point without overshoot under no-load conditions and under constant load conditions. The transition time is slightly longer than under no-load conditions, thus ensuring that the system has no overshoot under any circumstances and avoiding impact on the tool when transitioning from rapid traverse to constant load. 1 3) When the tool suddenly cuts into the workpiece and is loaded, as shown in Figure 4c, the system has no backward movement and no feed rate fluctuation, indicating that the system has sufficient rigidity. 4) The simulation curve shows that the rapid traverse speed of the system is approximately 140 mm/s. Therefore, in the machine tool cutting parameter table (omitted), the productivity of small-diameter pipe processing can be improved, while the productivity of large-diameter pipe processing is basically appropriate. 5) Under normal circumstances, this type of electro-hydraulic servo system can obtain good dynamic and static characteristics by using PID control. However, in applications where system parameters and external loads change significantly, using FC plus PID control can achieve better results. 6) The chamfering machine has been successfully debugged in a steel plant and used for nearly 2 years. In actual use, there is no continuous vibration (stable), no backing during loading (good rigidity), and no forward thrust during unloading (no tool breakage), thus meeting the design requirements. References : [1] Wang Zhanlin Modern Hydraulic Control [M]. Beijing: Machinery Industry Press, 1997. [2] Shi Weixiang et al. Modern Electromechanical Control Engineering [M]. Beijing: Machinery Industry Press, 1998.